In re Dealer Mgmt. Sys. Antitrust Litig.

Full title: IN RE DEALER MANAGEMENT SYSTEMS ANTITRUST LITIGATION, MDL 2817 This…

Court: United States District Court, Northern District of Illinois

Date published: Jun 29, 2023

680 F. Supp. 3d 919 (N.D. Ill. 2023)

Opinion

MDL 2817 No. 18 C 864

2023-06-29

IN RE DEALER MANAGEMENT SYSTEMS ANTITRUST LITIGATION, This document relates to: All Cases

REBECCA R. PALLMEYER, United States District Judge

MEMORANDUM OPINION AND ORDER

REBECCA R. PALLMEYER, United States District Judge

Plaintiffs in these consolidated cases—Loop, LLC ("AutoLoop"); a group of retail automobile dealerships (the "Dealers");¹ and Motor Vehicle Software Corporation ("MVSC")—have sued Defendants CDK Global, LLC ("CDK") and the Reynolds and Reynolds Company ("Reynolds") alleging antitrust violations. In short, Plaintiffs accuse Defendants of anticompetitive conduct in the market for "dealership management system" software and a related submarket for "data integration"—that is, a service that renders the data stored in those systems commercially functional. Both AutoLoop and the Dealers, who comprise a large group of named Plaintiffs, bring putative class claims: AutoLoop on behalf of a group of automotive software application vendors and the Dealers on behalf of a nationwide class of retail automobile dealerships. (See Dealers Compl. [198] at 1 (listing "Dealership Plaintiffs").) As discussed in greater detail below, this case originally included a fourth plaintiff—Authenticom, Inc. ("Authenticom")—which has since settled with both Defendants but whose allegations and legal arguments remain relevant to Defendants' pending motions.

1 Both AutoLoop and the Dealers bring putative class claims—AutoLoop on behalf of a group of automotive software application vendors and the Dealers on behalf of a nationwide class of retail automobile dealerships. The Dealers are represented by a large group of named plaintiffs. (See Dealers Compl. [198] at 1 (listing "Dealership Plaintiffs").)

This memorandum opinion and order resolves several pending motions. In summary, the court holds as follows:

• Defendants' motion to bar what they call a "new 'initial conspiracy' theory" offered by Plaintiffs' experts is denied.

• Reynolds's motion to bar MVSC's "newly disclosed claims" is denied.

• CDK's motion for summary judgment against AutoLoop is granted as to damages attributable Plaintiffs who have settled their claims ("Cox" Plaintiffs) and otherwise denied as to the Sherman Act § 1 conspiracy claim. CDK's motion is granted as to AutoLoop's Sherman Act § 1 exclusive-dealing claim. AutoLoop is directed to show cause why CDK's motion should not be granted as to AutoLoop's § 2 monopolization claim.

CDK's motion on AutoLoop's state law claims is denied to the extent those claims are based on the alleged conspiracy but otherwise granted. AutoLoop's claims will be subject to rule-of-reason analysis at trial.

• CDK's motion for summary judgment against the Dealers is granted on all Sherman Act damages claims but denied as to the § 1 conspiracy claim for injunctive relief. CDK's motion on the Dealer's state-law claims is granted as to Counts IX, XLIV, XXXIX, XL, XLVI, and XLVIII and is otherwise denied.

• Reynolds's motion for summary judgment against MVSC is granted as to any claim based on MVSC's failure to break into the Wisconsin market for electronic vehicle registration services and as to damages MVSC sustained after settling with CDK in October 2019. Reynolds's motion is also granted as to MVSC's California Unfair Competition Law claim. Reynolds's motion is otherwise denied. MVSC's claims will be subject to rule-of-reason analysis at trial.

INTRODUCTION

The court has detailed the basic facts of this MDL in multiple opinions and assumes knowledge of those opinions here.² Defendants CDK and Reynolds are automotive technology companies that offer "dealer management system" ("DMS") software—complex enterprise computer systems employed by car dealerships to collect, manage, and deploy data they generate.

2 See, e.g., In re Dealer Mgmt. Sys. Antitrust Litig. ("Authenticom MTD Op."), 313 F. Supp. 3d 931 (N.D. Ill. 2018); In re Dealer Mgmt. Sys. Antitrust Litig. ("AutoLoop MTD Op."), 362 F. Supp. 3d 477 (N.D.Ill. 2019); In re Dealer Mgmt. Sys. Antitrust Litig. ("Dealers MTD Op."), 362 F. Supp. 3d 510 (N.D. Ill. 2019); In re Dealer Mgmt. Sys. Antitrust Litig. ("Daubert Op."), 581 F. Supp. 3d 1029 (N.D. Ill. 2022).

Car dealers also use software applications ("apps") to provide services like inventory management, customer relationship management, warranty services, repair orders, and electronic vehicle registration and titling. Some apps are furnished by DMS providers—like Defendants—but most are created by third-party developers known as "vendors." To create useful apps for dealerships, vendors need access to the data located on dealers' DMSs. But data stored on the DMS is raw and unprocessed, making it difficult for vendors to use. So, vendors hire intermediaries—known as "data integrators"—who process, standardize, and export data from the DMS to vendors' platforms. In other words, data integrators (as their name suggests) "integrate" data stored in the DMS to make it usable by vendors' apps.

Reynolds and CDK themselves provide data integration services, which they sell to vendors; vendors can also purchase such services from unrelated third-party providers. Reynolds's and CDK's offerings are known as "certified" integration, which provide vendors with automated, secure access to DMS data for a fee.³ Third-party data integrators, like Authenticom, provide a similar service, which is known as "independent"—as opposed to "certified"—integration.⁴ CDK owns two such independent

3 CDK's certified integration product is called Third Party Access ("3PA"); Reynolds's is known as the Reynolds Certified Interface ("RCI").

4 The (ostensible) differences between certified and independent integration are discussed in detail later in this opinion.

integrators: Digital Motorworks, Inc. ("DMI") and IntegraLink, which compete with companies like Authenticom.

Historically, both CDK and Reynolds provided dealers with "open" DMSs, meaning that—explicitly or implicitly—CDK and Reynolds permitted dealers to give third parties direct access to DMS data. Over time, however, Reynolds began to "close" its DMS by selectively blocking independent data integrators from accessing dealer data stored on the Reynolds DMS. As Reynolds's blocking efforts reduced vendors' ability to use independent integrators on dealerships with a Reynolds DMS, vendors who needed to access data from a Reynolds DMS were forced to turn to Reynolds, which profited by increasing the fees it charged for certified data integration through its own RCI program. CDK's system, by contrast, remained "open" to third parties—a feature that CDK used to market its service to current and prospective customers.

Plaintiffs allege that by the summer of 2013, Reynolds was taking aggressive action to prevent independent integrators from serving dealers with a Reynolds DMS. CDK's subsidiaries—DMI and IntegraLink⁵—were, at that time, providing independent integration services to vendors who needed access to data stored on a Reynolds DMS. So Reynolds's "blocking" efforts began to undermine DMI's business. Then, in September 2013, Plaintiffs allege, CDK abruptly stopped emphasizing that its DMS was "open"—i.e., that vendors could, with dealer approval, access DMS data using an independent integrator. According to Plaintiffs, senior executives from CDK and Reynolds orally agreed at that time to restrict access to dealer data and to "destroy" independent data integrators like Authenticom and drive them from the market. Among other methods, CDK and Reynolds allegedly agreed to coordinate a message to the market: that, for "data security" reasons, they needed to block independent data integrators from accessing their respective DMSs. As CDK and Reynolds negotiated a formal written agreement along this line, CDK began planning its own effort to block independent integrators, while Reynolds significantly reduced the degree to which it had been blocking DMS access by DMI. In Plaintiffs' view, Defendants were acting according to a surreptitious agreement to help one another and suppress other competition.

5 Going forward, the court will refer to DMI and Integralink collectively as "DMI."

In February 2015—following their alleged oral agreement—CDK and Reynolds entered into three written agreements: (1) the Data Exchange Agreement or "Wind Down" Agreement; (2) the 3PA Agreement; and (3) the RCI Agreement (collectively, the "2015 Agreements"). In the Wind Down Agreement, CDK agreed to wind down its data integration business on the Reynolds DMS within approximately five years. In other words, CDK agreed that DMI—the independent integrator owned by CDK—would transition away from offering integration services for dealers who used a Reynolds DMS. In return, Reynolds agreed not to block DMI's access to the DMS during the wind-down period. As a result, DMI could continue to extract dealer data just as they had before, using login credentials provided by dealers. But CDK and Reynolds both agreed not to assist any other independent data integrator in any effort to access or integrate with the other Defendant's DMS. At the same time, the parties agreed to the 3PA Agreement and the RCI Agreement, which provided CDK and Reynolds with reciprocal access to each other's certified integration programs. Plaintiffs allege that shortly after CDK and Reynolds entered into the Wind Down Agreement, CDK began

renegotiating contracts with vendors for 3PA access. As part of that renegotiation, CDK required those vendors who used 3PA to access data on any CDK DMS to exclusively use 3PA for all customers with a CDK DMS.

By eliminating competition for data integration services, Plaintiffs claim that CDK and Reynolds have usurped control over dealer data and thwarted dealers' ability to control who can access and use their data. As a result, vendors have no choice but to access certified data integration from CDK and Reynolds, at a price much higher than they would pay to an independent integrator.

PROCEDURAL HISTORY

Dozens of parties, including independent data integrators and dealers, sued CDK and Reynolds for violations of the Sherman Act and a variety of state antitrust, consumer protection, and other laws. The Judicial Panel on Multidistrict Litigation ("JPML") consolidated the cases in this MDL. Some of the claims have been dismissed, while others have been resolved through settlement.⁶ For purposes of this opinion, the governing complaints include: (1) AutoLoop's amended complaint against CDK and Reynolds [194]; (2) the proposed Dealer class's consolidated complaint against CDK [198]; and (3) MVSC's second amended complaint against Reynolds, see Second Amended Complaint [76], Motor Vehicle Software Corp. v. CDK Global, Inc. et al., No. 2:17-cv-896 (C.D. Cal. Nov. 2, 2017). Not all of the claims asserted in those complaints are still live, however; the claims that remain active include: AutoLoop's claims against CDK for horizontal conspiracy and exclusive dealing in violation of Sherman Act § 1, monopolization in violation of Sherman Act § 2, and parallel Florida state law claims; the proposed Dealership Class's claims against CDK for conspiracy and exclusive dealing in violation of Sherman Act §§ 1 and 2 and related state antitrust and consumer protection laws; MVSC's claims against Reynolds for conspiracy in violation of Sherman Act §§ 1 and 2 and related California and Illinois state law laws; and Reynolds and CDK's counterclaims against various Plaintiffs.⁷

6 The Dealers and Reynolds entered into a settlement agreement in January 2019. (See [427], [502].) MVSC settled its claims against CDK and CVR in October 2019. (See [778].) Authenticom and CDK settled in October 2020 (see [1199]), and Authenticom and Reynolds settled in May 2022 (see [1351]).

7 The court resolves counterclaims-related motions in separate, simultaneously-released memorandum opinions and orders.

The court's most recent memorandum opinion and order resolved the parties' Daubert motions. See Daubert Op., 581 F. Supp. 3d at 1042. In that ruling, the court granted in part and denied in part the motions to exclude Dr. Michael A. Williams and Allan Stejskal, and denied the motions to exclude Edward M. Stroz, Gordon Klein, Daniel L. Rubinfeld, Dr. Kevin Murphy, Dr. Mark Israel, Catharine Lawton, Nancy Miracle, and Brian Halpin.⁸ Following its Daubert ruling, the court invited the parties to file supplemental briefs on the impact of those rulings on

8 Dr. Israel serves as an expert for Authenticom, AutoLoop, and MVSC as to antitrust collusion and damages. Dr. Williams is as an expert for the Dealers on those same issues. Dr. Murphy is CDK's rebuttal expert as to those same subjects. Lawton serves as an expert for Authenticom on lost profits, and Rubinfeld is Defendants' rebuttal expert on that issue. Klein serves as MVSC's expert on lost profits. Stejskal is an expert on the automotive industry for all Plaintiffs. Stroz serves CDK's expert on cybersecurity. Halpin is as an expert for Plaintiffs on authentication issues related to Stephen Cottrell's notes on a conversation with Dan McCray from CDK, which is discussed at length below.

the pending summary judgment motions. (See [1322].) Those supplemental briefs are now before the court. (See [1327]-[1334].) This memorandum opinion and order resolves the pending motions for summary judgment ([954], [967], [970]), as well as Defendants' motion to bar certain claims and theories that, according to Defendants, were not timely disclosed. (See [773], [787].) The court turns, first, to the motions to bar, because their resolution determines the scope of the summary judgment issues remaining to be decided.

PRELIMINARY MOTIONS

Defendants have moved to bar theories that, according to Defendants, Plaintiffs disclosed for the first time in expert reports. As Plaintiffs never sought to amend their complaints to include the "new" theories, Defendants urge, the court should bar Plaintiffs from relying on those theories in opposing summary judgment.

I. Legal Standard

Defendants do not identify a particular rule under which the court should evaluate their motions, citing instead to out-of-circuit cases decided in different procedural postures. (See Defs.' Mot. to Bar Initial Conspiracy Theory [775] at 11.) Defendants' argument is best understood as a challenge to Plaintiffs' asserting new arguments in summary judgment briefing. In assessing such a challenge, the Seventh Circuit has instructed that the "first step" is to consider whether Plaintiffs' "new" arguments "change[] the complaint's factual theory" or whether they change legal theories.⁹ BRC Rubber & Plastics, Inc. v. Cont'l Carbon Co., 900 F.3d 529, 540 (7th Cir. 2018). If the factual theory has changed, "the plaintiff may be attempting in effect to amend its complaint, and the district court has discretion to deny the de facto amendment." Id. Leave to amend is "freely give[n] when justice so requires," FED. R. CIV. P. 15(a)(2), but may be denied "where there is a good reason to do so: 'futility, undue delay, prejudice, or bad faith.'" R3 Composites Corp. v. G&S Sales Corp., 960 F.3d 935, 946 (7th Cir. 2020) (quotation omitted). In general, "[a]n amended pleading is less likely to cause prejudice if it comes without delay or asserts claims related to allegations asserted in prior pleadings," while "prejudice is more likely when an amendment comes late in the litigation and will drive the proceedings in a new direction," Allen v. Brown Advisory, LLC, 41 F.4th 843, 853 (7th Cir. 2022), forcing "parties to re-litigate issues that have already been settled," "incur additional discovery costs," or have their "litigation preparation wasted." Hoenig v. Karl Knauz Motors, Inc., 983 F. Supp. 2d 952, 960 (N.D. Ill. 2013).

9 Defendants' motions focus only on (supposedly) new factual theories, not legal theories.

II. Defendants' Motion to Bar Plaintiffs' "New 'Initial Conspiracy'" Theory

In their first motion, Defendants object to expert opinions that were disclosed four months after the close of expert discovery in April 2019: opinions offered by Dr. Mark Israel (who submitted a report on behalf of Authenticom, AutoLoop, and MVSC), Dr. Michael Williams (who submitted a report on behalf of the Dealers), and Catharine Lawton (who submitted a report on behalf of Authenticom). Defendants contend that those reports would "fundamentally amend [Plaintiffs'] complaints" by asserting "for the first time a new 'Initial Conspiracy' starting in 2013, separate from the alleged 2015 conspiracy that Plaintiffs have maintained since the outset of the litigation." (Defs.' Mot. to Bar Initial Conspiracy Theory at 1.)

According to Defendants, Plaintiffs' complaints sound the "constant and unmistakable

theme" that Defendants (1) "historically competed with one another based on the extent to which each allowed third-party access to their respective dealer management systems," but (2) "'abruptly' stopped competing in February 2015, when CDK 'began blocking dealers from granting third-party access to dealer data and, at the same time, agreed to shut down its data integration business for dealers using a Reynolds DMS.'" (Id. (quoting AutoLoop Compl. [194] ¶ 16).) Now—in contrast, Defendants contend—Dr. Israel dates the alleged collusion as beginning as early as September 2013, when Reynolds allegedly "took aggressive action to block" independent integrators and raise integration prices and "worked with CDK to avoid blocking CDK's DMI integration service on Reynolds's system." (Id. at 6 (quoting Israel Rep. [889-1]).) Williams and Lawton also peg September 2013 as the start of the conspiracy. Defendants assert that this has added "perhaps hundreds of millions of dollars in new antitrust damages" to the litigation and argue that it would be "enormously prejudicial" to allow Plaintiffs to amend their complaints now "well after substantial motions practice, two years of litigation, and the close of fact discovery." (Id. at 1.)

The issue here is whether Plaintiffs' new "initial conspiracy" allegations alter the complaints' original factual theory. BRC Rubber, 900 F.3d at 540. Defendants cite the complaints' "many references to 2015 as the fulcrum point of Plaintiffs' claims" and recount several earlier rulings in this case in which the 2015 Agreements were viewed as the "focal point" of the lawsuit. (Defs.' Mot. to Bar Initial Conspiracy Theory at 4, 12-13.) The court does not believe Plaintiffs' allegations (or the court's understanding of them) are so rigidly limited. True, as Defendants emphasize, the complaints focus on the 2015 Agreements. But Plaintiffs alleged collusive conduct that occurred earlier. They pleaded, for example, that "[t]op executives at CDK and Reynolds discussed blocking Authenticom at least as early as June 2014." (See AutoLoop Compl. ¶ 93.) They also cited internal CDK documents from later in 2014, which reference a need to "enter into 'reciprocal agreements with DMS providers' including specifically Reynolds, in order to eliminate competing data integrators." (Id. ¶ 94.)

In denying CDK's motion to dismiss AutoLoop's complaint, the court recognized that the alleged conspiracy was "not limited to" the 2015 Agreements. AutoLoop MTD Op., 362 F. Supp. 3d at 491. The court's ruling discussed alleged agreements between Defendants "'to block and thereby destroy third-party data integrator[s]'" and not to "'compete with each other for data integration services.'" Id. (quoting AutoLoop Compl.). Though the complaint primarily cited evidence post-dating the 2015 Agreements (but see AutoLoop Compl. ¶¶ 93-94), Plaintiffs cite evidence produced in discovery showing that Defendants' collusion began in September 2013, when negotiations for the 2015 Agreements began. Discovery also revealed additional evidence about how CDK allegedly "supported" Reynolds's efforts to block independent integrators prior to February 2015. The court reviews that evidence in detail in its discussion of the merits.

This is not a "new" factual theory that requires Plaintiffs to formally amend their complaints. In relying on evidence obtained in discovery to flesh out the contours of Defendants' alleged conspiracy to "close" their DMS, Plaintiffs have not altered the "fundamental factual allegation" in their complaints. Compare Whitaker v. Milwaukee Cnty., 772 F.3d 802, 807 (7th Cir. 2014) (plaintiff could switch from "agency" theory to "joint employer" theory where change did not alter complaint's

"fundamental factual allegation") with Chessie Logs. Co. v. Krinos Holdings, Inc., 867 F.3d 852, 860-61 (7th Cir. 2017) (declining to consider a new theory when plaintiff originally alleged that defendant dumped dirt on plaintiff's property but argued on summary judgment that defendant removed dirt from plaintiff's property). In any event, if the court were to conclude that the "initial conspiracy period" allegations in fact represent a change in the complaints' factual theory, the court would be inclined to grant leave to amend, and allow allegations based on information obtained for the first time in discovery. Plaintiffs' purportedly new "initial conspiracy" theory is, at a minimum, "related to allegations" in the governing complaints. See Allen, 41 F.4th at 853.

Defendants have not demonstrated prejudice. The challenged theories do not "drive the proceedings in a new direction," and Defendants' experts had an opportunity to rebut them: Plaintiffs' three experts were deposed extensively, and all three were subjected to Daubert challenges. Defendants have by now moved for summary judgment on the "new" theories, but insist that, had they known Plaintiffs' "new" theory, they would have demanded a "host of additional discovery from vendors, dealers, and potentially car manufacturers about the state of the marketplace in 2013 and 2014" to "establish Reynolds's unilateral and legitimate reasons for its business decisions regarding third-party access as well as general market conditions." (Defs.' Mot. to Bar Initial Conspiracy Theory at 17.) Defendants also claim that they would have "sought documents and financial data—from Authenticom, Dealers, and third parties—back to at least 2011, rather than January 1, 2013." (Id. at 18.) But Defendants have not identified any specific material that could have been revealed, had they explored those topics, and have not asked the court to reopen discovery on these topics. Nor have Defendants argued, In their extensive briefs, that they lack evidence to show they are entitled to summary judgment in their favor. Indeed, several of the categories of documents that Defendants claim they would have requested call for information in Defendants' own possession. Those categories include documents that would answer questions such as: "Did Reynolds also 'avoid blocking' Authenticom and the other so-called integrators in 2014?" and "Did Reynolds roll out any security measures during that period that hampered third-party access to the Reynolds DMS?" (Id. at 17.)

Finally, to the extent that Defendants had to "incur additional discovery costs" to defend against the "initial conspiracy" allegations, Hoenig, 983 F. Supp. 2d at 960, those costs have already been incurred: Defendants' experts prepared rebuttals, everyone has been deposed, the parties briefed Daubert and summary judgment, and the court has expended significant resources resolving those motions. The court declines to depart from the general principle that disputes should be resolved "on the merits rather than [on] technicalities." Stanard v. Nygren, 658 F.3d 792, 800-801 (7th Cir. 2011).

III. Reynolds's Motion to Bar MVSC's "Newly Disclosed Claims"

MVSC—a company in the electronic vehicle registration ("EVR") business—alleges that Reynolds and CDK conspired to prevent it from using either independent integrators, like Authenticom, or certified integration through RCI or 3PA. In a second motion to bar, Reynolds argues that MVSC is improperly expanding its claim for damages in the expert report submitted by Gordon Klein that "include[s] new damages claims based on new markets: the Virginia and Wisconsin EVR markets." (Reynolds Mot. to Bar Newly Disclosed Claims [789] at 1.) As Reynolds

reads MVSC's complaint, it "sound[s] a constant and unmistakable theme" focused on allegedly unlawful conduct by Reynolds in "two specific state EVR markets: California and Illinois." (Id.) Reynolds emphasizes that "[o]ut of 263 paragraphs in the [second amended complaint], only two of them mention the Virginia and Wisconsin EVR markets, briefly noting that MVSC does not compete in those markets." (Id. at 3.) And according to Reynolds, the complaint's "paragraphs constituting the causes of action exclusively refer to alleged harm in the Illinois and California EVR markets." (Id. at 7.) Reynolds contends the court should bar MVSC from pursuing claims based on harm allegedly suffered in the Virginia or Wisconsin EVR markets and strike such opinions from Klein's report. Again, the court disagrees.

Reynolds's claim of surprise arising Klein's report ignores important allegations in the governing complaint. MVSC alleged that Reynolds and CDK conspired to "(1) protect CVR's¹⁰ monopoly in the Illinois EVR market; (2) empower CVR to regain its monopoly position in the California EVR market; and (3) block MVSC from effectively competing with CVR in markets that MVSC may enter in the future." (MVSC Compl. ¶ 107 (emphasis added).) In a section on "Other State EVR Markets," MVSC alleges: "MVSC plans to enter other EVR markets, including Wisconsin and Virginia. But so long as CDK and Reynolds deny MVSC access to the dealer data stored in their DMS platforms, it will be very difficult, if not impossible, for MVSC to compete in those markets." (MVSC Compl. ¶ 86.) MVSC's § 1 claim does refer specifically to California and Illinois. But it also incorporates by reference the complaint's prior allegations, including those quoted above.

10 CVR—a shorthand for "Computerized Vehicle Registration"—is a joint venture between Reynolds and CDK.

As MVSC notes, some discovery correspondence casts doubt on Reynolds's claim of surprise. In a letter to MVSC, cc'd to Reynolds's attorneys, counsel for CDK and CVR wrote: "You [MVSC] accepted our proposal to limit the production ... to documents for California, Illinois, Virginia, and Wisconsin" for certain document requests, because "according to MVSC's Second Amended Complaint, those are the only states in which MVSC allegedly operates or is working to gain approval to operate." ([818-4] at 3-4.) Reynolds did not write the letter, but it would have put them on notice that MVSC's claims were not limited to California and Illinois.

Finally, Reynolds has not shown that it will suffer prejudice if its motion to bar is denied. For reasons explained below, the court is granting summary judgment against MVSC with respect to Wisconsin-based damages because the evidence does not support a claim that the alleged conspiracy resulted in those damages. See Section III.B.3.c, infra. With respect to damages arising in the Virginia market, Reynolds's claim of prejudice is defeated by documents that put Reynolds on notice that MVSC's claim involved Virginia. Further, MVSC's Rule 30(b)(6) witness, CEO Don Armstrong, testified about the harm MVSC suffered in its attempted expansion into Virginia. (See MVSC Mot. to Bar Opp. [818] at 15.) While Reynolds claims that it would have requested other documents and "asked many more questions" when it deposed Armstrong and MVSC COO Joe Nemelka (Reynolds Mot. to Bar Newly Disclosed Claims at 11), it does not identify any specific Virginia-related information it was unable to obtain due to MVSC's allegedly "late" disclosure. Because MVSC sufficiently

established in its complaint and discovery that it was claiming Virginia-related damages—and Reynolds is not prejudiced in any event—the court denies Reynolds's motion to bar.

SUMMARY JUDGMENT MOTIONS

I. Legal Standard

The summary judgment standards are familiar. The court will grant such a motion where "the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." FED. R. CIV. P. 56(a); see Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). A genuine issue of material fact exists if "the evidence is such that a reasonable jury could return a verdict for the nonmoving party." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). The court draws reasonable inferences in favor of the nonmoving party, but "may not make credibility determinations, weigh the evidence, or decide which inferences to draw from the facts; these are jobs for a factfinder." Johnson v. Rimmer, 936 F.3d 695, 705 (7th Cir. 2019) (quotation omitted). A party opposing summary judgment must go beyond the pleadings and "set forth specific facts showing that there is a genuine issue for trial." Liberty Lobby, 477 U.S. at 250, 106 S.Ct. 2505. Summary judgment is proper if the nonmoving party "fails to make a showing sufficient to establish the existence of an element essential to that party's case, and on which that party will bear the burden of proof at trial." Ellis v. CCA of Tenn. LLC, 650 F.3d 640, 646 (7th Cir. 2011) (quoting Celotex, 477 U.S. at 322, 106 S.Ct. 2548).

The court considers "only those facts whose substance would be admissible at trial under a form permitted by the Federal Rules of Evidence, although the form produced at summary judgment need not be admissible." Wragg v. Vill. of Thornton, 604 F.3d 464, 466 (7th Cir. 2010). If a "non-moving party objects to evidence cited in the moving party's motion for summary judgment, the court considers that evidence only if the court concludes it would be admissible at trial." Allied Ins. Co. v. United States, 627 F.Supp.3d 944, 955 (N.D. Ill. 2022) (citation omitted).

II. CDK's Motions for Summary Judgment on Plaintiffs' Claims

The court turns, first, to CDK's motion for summary judgment against AutoLoop [969] ("CDK AutoLoop MSJ") and CDK's motion for summary judgment against the Dealers [973] ("CDK Dealers MSJ"). Because both parties incorporated the Authenticom briefing, the court considers arguments made in CDK's motion for summary judgment against Authenticom [966] ("Defs.' Authenticom MSJ") as well.¹¹

11 For ease of reference, the court attributes arguments made in the Authenticom briefing to the party adopting them. For example, the court may discuss "AutoLoop's argument" though the argument is found in Authenticom's opposition.

In summary: CDK argues that the evidence is at least as consistent with CDK acting in its own unilateral interest as it is indicative of an antitrust conspiracy. According to CDK, it had an independent profit motive for "closing" its DMS to independent integrators, irrespective of what Reynolds had done or planned to do. CDK also attacks various pieces of Plaintiffs' theory—CDK says it had no motive, the agreement had no enforcement mechanism, and that the data on dealerships that switched DMS providers is inconsistent with reduced competition. On Plaintiffs'

unilateral Sherman Act claims, CDK urges that Plaintiffs have not identified a causal link between the supposed "exclusive dealing" provisions and Plaintiffs' damages. Finally, CDK asserts several state-specific arguments based on issues like nexus, standing, and damages.

A. Background

The following facts are taken primarily from the parties' Local Rule 56.1 statements and are undisputed unless otherwise noted. The court also draws from its prior opinions and the parties' briefs for background information that may help the reader.

1. The DMS market

Auto dealers use DMS software to manage business functions including accounting, sales, inventory, service, customer information, and human resources. Dealers typically store customer names, addresses, emails, and telephone numbers in their DMS and may store Personally Identifiable Information ("PII") and Sensitive Personal Information ("SPI"), like Social Security numbers, driver's license numbers, and financial information. (See Pls.' Resp. to Defs.' Joint Statement of Undisputed Material Facts [1157] ("PRDSOF") ¶ 52.) The DMS may also include intellectual property and proprietary data that belongs to the DMS provider, to Original Equipment Manufacturers ("OEMs"), or to other third parties. (Id. ¶ 54.)

CDK and Reynolds are the two largest DMS providers, who together control 70 percent of the U.S. franchise dealership market. During the relevant period, CDK and Reynolds had customers in all fifty states. (See Defs.' Resp. to Pls.' Joint Statement of Undisputed Material Facts [1125] ("DRPSOF") ¶¶ 130, 136-37.)

Reynolds developed its ERA DMS¹² in the late 1980s. (PRDSOF ¶ 2.) CDK was formed in 2014 when Automatic Data Processing, Inc. ("ADP") spun off its Dealer Services division into a stand-alone company, CDK.¹³ (See id. ¶ 16.) Both Reynolds and CDK license their DMS software to dealers. (See id. ¶ 65.) The term of Reynolds's DMS contract ranges from three to seven years; the typical term is five years. (Id. ¶ 4.) By its terms, the Reynolds Master Agreement prohibits licensees from providing DMS access "to any third party, except your employees who have a need for access to operate your business and who agree to comply" with the dealer's contractual obligations. (Id. ¶ 66.) As Plaintiffs note, however, the agreement defines "You" to mean dealers and their "employees, agents, and representatives." (Id.)

12 The acronyms at play in this case are many and varied. The court is uncertain of the origin or significance of "ERA" for Reynolds' DMS.

13 For ease of comprehension, the court will refer to the entity as "CDK," even when discussing actions prior to the 2014 spinoff, unless the distinction between CDK and ADP is contextually significant.

Most CDK DMS contracts are three to five years long. (Id. ¶ 18.) CDK's standard Master Services Agreement prohibits access "by any third parties except as otherwise permitted by this agreement." (Id. ¶ 67.) Some CDK DMS contracts, however, added non-standard addenda in which CDK acknowledged the dealer's right to provide a third party with a user ID and password that permitted data extraction. (Id. ¶ 69.) Even apart from these addenda, Plaintiffs note, the CDK agreement permits dealers to provide DMS access to "agents"; Plaintiffs thus dispute that the addenda were necessary for dealers to authorize access by agents. (See id. ¶¶ 67, 69.)

DMS providers seeking to recruit new customers face challenges, though the parties

dispute the significance of those challenges. (DRPSOF ¶ 1.) Plaintiffs cite various obstacles that inhibit dealers who might want to switch DMS providers: long contract terms, substantial early termination fees, auto-renewal provisions, and out-of-pocket setup expenses and training fees. (Id. ¶¶ 2, 4.) Dealers opting to switch from one DMS provider to another also risk losing valuable historical data. (Id. ¶ 3.) Defendants generally admit these obstacles and costs, but they note testimony of Plaintiffs' expert Allan Stejskal that, for "many years, DMS providers have subsidized the costs of switching," for example by offering discounts and free training. (Id. ¶¶ 2, 4; PRDSOF ¶ 175)

Some large dealerships have abandoned attempts to switch DMS providers, at least in part because of high switching costs. (See DRPSOF ¶ 7.) The parties dispute how often dealers switch DMS providers at all. (Id. ¶ 8.) According to Plaintiffs, dealers on average stay with CDK or Reynolds for approximately 20 years, and no more than 5 percent of dealerships change DMS providers in any given year. (Id. (citing 2014 CDK presentation stating that "clients have an implied average tenure of 20 years" and Dr. Israel's expert report, which calculates a 4.4 percent switching rate in 2017).) Reynolds Vice President Bob Schaefer declared that, as of 2017, the average dealership tenure was "less than 20 years" and "shrinking," but did not specify a length of time. (Id.)

2. The application and data integration markets

In addition to their DMS, dealerships typically use 15 to 20 software applications ("apps") to manage their retail operations. (DRPSOF ¶ 11.) CDK and Reynolds offer their own applications (sometimes called "layered applications"), but they are not market leaders among providers of these apps. (Id. ¶ 12.) Instead, dealers typically engage third-party software application providers—referred to as "vendors"—to provide services like inventory management, customer relationship management, warranty services, repair orders, and electronic vehicle registration and titling. See generally Daubert Op., 581 F. Supp. 3d at 1042. AutoLoop, which sells software to thousands of automotive dealerships, is one such vendor. AutoLoop offers three software "suites"—for sales, service, and marketing—each of which includes multiple software applications. (AutoLoop Compl. ¶¶ 6-7.)

To provide services to a dealer, vendors need to access data located on the dealer's DMS. Vendors typically do so by purchasing data integration services from data integrators, who collect and standardize DMS data, making it usable for vendors. Integrators also sometimes provide additional services for the vendors, for example, standardizing data from different DMSs or correcting data errors. See generally Daubert Op., 581 F. Supp. 3d at 1042. To function optimally, some apps need real-time, event-driven integration—in other words, integration in which updates in the DMS automatically, and near-instantly, update in the app. (PRDSOF ¶ 32.)

CDK and Reynolds each offer their own data integration service for their DMSs—"3PA" and "RCI," respectively—which serve as alternatives to the purchase of services from data integrators. Historically, many third-party data integrators—including Authenticom—offered relatively cheap data extraction from DMSs. (DRPSOF ¶¶ 13-14.) Plaintiffs offer evidence that these independent integrators, by offering cheaper access to dealer data, make it easier for startup software vendors to enter the market. (Id. ¶ 15.) Defendants note that independent integrators—whom they call "hostile data extractors"—are not "the only or the best solution for

startup software vendors to enter the market," but Defendants do not meaningfully contest that the independent extractors are cheaper. (Id.)

a. Reynolds's RCI program

Reynolds created RCI in 2001 to manage third-party integration with its DMS. (PRDSOF ¶ 6.) To join the RCI program, vendors must sign a license agreement, which is mutually terminable at will, typically on 180 days' notice or less. (Id. ¶¶ 7, 10.)

RCI customizes its interface to vendors' needs and offers real-time update triggers—meaning the vendor is automatically updated whenever a particular data field is changed in the DMS—and writeback capability (i.e., the ability to "write" or input data back into the DMS). (Id. ¶ 8.) Reynolds does not restrict vendors who purchase integration services from RCI from receiving or using data that a dealer itself downloads and exports—i.e., sends to the vendor—from the Reynolds DMS. (Id. ¶ 9.)

b. CDK's 3PA program

CDK created its certified data integration service, 3PA, in 2000. Vendors in the 3PA program sign a standard agreement with CDK called a Third-Party Access Agreement. (PRDSOF ¶ 24.) CDK maintains that since at least 2006, vendors using 3PA have been obligated to obtain data within the CDK DMS only via the 3PA program, and not from any other source. (Id.; see also DRPSOF ¶ 67.) AutoLoop disputes that this has been the case since 2006 and contends, to the contrary, that "those restrictions have materially changed since 2006." (PRDSOF ¶ 24.)

It is undisputed that 3PA vendors can pay to receive "bi-directional" or "writeback" integration for certain data. (Id. ¶ 25.) Plaintiffs respond that such integration is very expensive and dispute that CDK provides the data "necessary for the application" through 3PA and claim that CDK denies 3PA vendors important data functionality. (See id.)

CDK presents evidence of vendors who have acknowledged the benefits of "direct" integration through 3PA. Other have been more critical. For instance, plaintiff vendor AutoLoop cites testimony that 3PA's services were no better than the real-time, bi-directional integration AutoLoop received from an independent integrator called SIS. AutoLoop claims it received more functionality from SIS at a fraction of the price. (See id. ¶ 26.)

c. CDK's DMI and IntegraLink

In addition to its certified integration service (3PA) for its own DMS, CDK acquired two companies that offered data integration services for the Reynolds DMS: DMI in 2002 and IntegraLink in 2010. (PRDSOF ¶¶ 27-28.) From 2010 to 2012, DMI's revenues nearly doubled—from $43 million to $70 million—as did its profits. (DRPSOF ¶ 31.) As discussed in greater detail below, DMI's continued third-party access to the Reynolds DMS became a core piece of the 2015 Agreements reached between CDK and Reynolds.

d. Authenticom

Authenticom is an independent data integrator. Plaintiffs cite declarations and witness testimony—which Defendants dispute with opposing declarations and witness testimony—that Authenticom makes data easier for vendors to use and provides dealers with greater transparency than 3PA or RCI. (See DRPSOF ¶ 18.) Vendors typically pay for Authenticom's services, but the parties dispute who "hires" Authenticom and therefore whether dealers have authorized Authenticom to access their DMS data. (PRDSOF ¶ 30.) Defendants claim that vendors are the parties who hire and pay Authenticom for integration services; Plaintiffs urge that, because dealers have sole authority to authorize

Authenticom and provide access to their DMS, vendors cannot "hire" Authenticom to perform data integration services without dealer authorization and cooperation. (Id.)

During the relevant period, Authenticom's standard service was once-daily "polling"—in other words, once a day, Authenticom would pull the previous 24 hours' data from the DMS. (Id. ¶ 31.) Authenticom has never licensed Reynolds or CDK DMS software—in other words, entered into a license agreement and paid for use of either company's software—though it did license "certified" data access from CDK between approximately 2008 and 2011. (Id. ¶ 79.) Plaintiffs note that dealerships are the direct DMS customer and their right to use the DMS software and database includes use by dealers' agents, including Authenticom. (See id.)

To access Reynolds DMS data, Authenticom used dealer-provided login credentials and ran Reynolds's DMS in the same way that a dealership employee would. (Id. ¶ 73.) To access CDK's DMS, Authenticom establishes a remote connection to the dealer's computer system and logs in, using credentials obtained from the dealer. (Id. ¶ 77.) Authenticom uses queries to extract data from CDK DMS; the results are displayed on screen and then "scraped" from the screen and stored by Authenticom. (Id. ¶ 78.)

Defendants maintain that Authenticom has not been able to offer real-time, event-driven integration. (Id. ¶ 33.) Plaintiffs dispute this; they assert that Authenticom has offered "near real-time" services like 3PA, which updates every 15 minutes, and that Authenticom has the technological capability to provide real-time integration but has not marketed that offering because of CDK and Reynolds's blocking activities, which make real-time integration ineffective. (See id.) Defendants also maintain that Authenticom has never offered real-time writeback services, but Plaintiffs dispute this as well, identifying several clients for whom Authenticom provided those services. (Id. ¶ 34.)

3. Defendants' security and performance concerns

Citing security and performance concerns, Reynolds and CDK have attempted over the years to prevent what they contend is unauthorized DMS access by third parties. Plaintiffs dispute the legitimacy of these concerns as well as their factual basis. The parties agree that automated access by data integrators puts some processing burden on the DMS, but Plaintiffs dispute that Authenticom or other independent integrators impede system performance. (PRDSOF ¶¶ 43-44.)¹⁴

14 In approximately 2010, Authenticom caused a performance problem on the Reynolds DMS because of a malfunctioning query, which Authenticom's CEO, Steve Cottrell, and Reynolds Vice President Bob Schaefer worked together to resolve within the same day. (See PRDSOF ¶ 48.)

The parties also agree that data corruption issues can result when a third-party writes data back onto the DMS, but they disagree about whether data corruption is a significant problem and whether writeback via "non-certified interfaces" presents greater risks than writeback attempts via certified interfaces. (See id. ¶ 51.) The parties also dispute whether independent integrators can access consumer PII and/or SPI maintained in the DMS. Plaintiffs maintain that independent integrators do not have access to sensitive data like Social Security and credit card numbers. (See id. ¶ 53.)

There is no dispute that data security is important to the automotive industry, including dealers and data integrators. (See id. ¶ 61.) In August 2013, for example, the National Automobile Dealers Association

("NADA") issued a "Dealer Data Guidance" memo to dealers; the memo warned dealers that their collection of sensitive information subjects them to federal privacy laws and raises "difficult and sensitive issues for dealers that they must understand when they enter into contracts that could allow access" to dealer data. (Id. ¶ 63.) The memo advised dealers to limit third-party access to data "by gathering it, and sending it to [third-parties], rather than allowing them to 'take' it by accessing your systems." (Id.) Plaintiffs observe that NADA did not instruct dealers to avoid giving vendors DMS access; to the contrary, the guidance specifically contemplates that vendors will access the DMS and advises dealers to "conduct due diligence when selecting a service provider" and to "audit and monitor their access." (Id.)

A follow-up NADA memo in 2014 encouraged dealers to "no longer allow vendors to access your systems directly for any reason." (Id. ¶ 64.) Plaintiffs dispute, however, that this advice relates to data integration; they point instead to testimony from NADA's corporate representative that NADA's view was that "the dealer should have control over the data in their systems and be able to share it with whom they'd like." (Id.)

4. CDK's treatment of third-party integrators prior to the alleged conspiracy period (September 2013)

CDK historically supported dealers' right to use data integrators. (See DRPSOF ¶ 22 (citing records stating CDK's support for "dealer permissioned data integration").) Beginning in 2007, CDK offered "Basic Access," which permitted dealers to "provide[] access to third-part[ies] through dealer supplied User IDs and strong passwords as well as dealer provided modem access" while "maintain[ing] all responsibility for system and data access." (PRDSOF ¶ 140.) CDK claims it dropped Basic Access in October 2012, but Plaintiffs note evidence that in June 2013 CDK made a presentation discussing "Basic Access" for vendors. (Id.) Plaintiffs also generally respond that CDK continued to advise dealers they were free to grant access to third parties as late as 2015, even if the "Basic Access" moniker was no longer used. (See id.)

Between 2006 and 2013, CDK overtook Reynolds as the nation's largest DMS provider—a success various CDK and dealer witnesses attribute at least in part to CDK's "open" strategy. (See DRPSOF ¶ 23.) At his deposition, Reynolds's CEO and Chairman Bob Brockman acknowledged that CDK had been "harmful to [Reynolds] in the marketplace" and cost Reynolds hundreds of millions of dollars in lost revenues by "badmouth[ing]" Reynolds's more-restrictive "data access policies" "high and wide ... in the market." (Id. ¶ 25.)

In 2007, CDK's DMS was maintained on-site by dealers; third parties connected through dealer-provided modems. (PRDSOF ¶ 141.) The parties dispute whether CDK had the technical capability to prevent third-party access at that time. (See id.) In 2008 and 2009, CDK began migrating its DMS software from being maintained on-site by dealers to a "hosted" environment on CDK servers. (Id. ¶ 142.) CDK CEO Steve Anenen testified that CDK became more concerned about data security as it transitioned to hosted servers. But Plaintiffs dispute the truth of that testimony and note that CDK continued to promote dealers' ability to use data integrators on the CDK DMS for several more years. (See id.)

The date on which CDK changed course and began to block access by independent integrators is also disputed. Defendants maintain that it was 2010; Plaintiffs point to late August 2013. Specifically, according

to CDK, it began creating "profiles" for third parties that were accessing the DMS outside the 3PA program in 2010 and began to identify and catalogue third-party code installed on its DMSs. (Id. ¶ 138.) In July 2011, CDK identified such code on 2,017 out of 3,548 DMS servers—an "infection rate" of 56.8 percent. (Id. ¶ 139.) Defendants call third-party code "hostile," but Plaintiffs challenge that characterization; they point to testimony that independent integrators installed code only with dealer authorization and that Authenticom never installed its own software on a dealer's DMS. (See id.)

In support of their position that CDK's efforts to block access did not begin until August 2013, Plaintiffs point to CDK's March 2013 Information Security Guidelines. As CDK now highlights, those Guidelines advised dealers to "[n]ever allow third parties to use your user ID or 'screen-scrape' your data." (Id. ¶ 144.) But Plaintiffs emphasize additional context. The preceding paragraph of the guidelines stated: "Suppose there is a breach and you have a forensics team trying to find out where it came from. Ah ha — it was you! Third parties should always have unique IDs and be required to authenticate to the system before they can access the data. These credentials should be carefully managed." (Id.) Thus, Plaintiffs urge, the guidelines recognize that third parties will have access to data and do not suggest that dealers should avoid creating User IDs for third parties altogether.

In June 2013 (prior to ADP's spinoff of its business to form CDK), ADP Chief Business Security Officer Jim Foote wrote: "the amount of 3rd parties who put unregistered code on the DMS is troubling to say the least." (Id. ¶ 145.) He referred to the "painful[] aware[ness]" of "the amount of 3rd parties who put unregistered code on the DMS" and asked for a briefing on "the different avenues these hostile third parties use to exploit the DMS." (Id.) Plaintiffs admit that Foote made these statements, but dispute that CDK in fact considered the code "troubling" or a security threat. In support, Plaintiffs note that CDK knew about the unregistered code for more than five years before Foote's statements and that Foote's email never calls the code a "security threat." (See id.) In August 2013, CDK issued an internal report which found that every CDK DMS server that the report analyzed contained some level of data corruption, which was "a serious problem" for CDK and its clients and "a time bomb just sitting and waiting to go off ... unless we address the root causes." (Id. ¶ 146.) The report provided examples of data corruption issues attributed to "[h]ostile 3rd parties." (Id.) To address data corruption problems, the report recommended that CDK "[c]ontinue to tighten system security and the ability to alter data without using ADP approved interface[s]" and eliminate "hostile interfaces." (Id.) Plaintiffs dispute that the report attributes data corruption solely to "hostile" integrators and emphasize that technical errors by CDK and manual data entry by dealers also contribute to data corruption. (Id.)

On August 28, 2013, Greg Meyer (ADP's CTO) emailed Steve Anenen (then the President of ADP Dealer Services and CDK's future CEO) and Bob Karp (then a Senior Vice President at CDP Dealer Services, and CDK's future President of CDK North America) to inform them that he had "started the conversation regarding securing our DMS from hostile 3rd party access. I will kick-off a project tomorrow to scope the effort and provide weekly updates." (Id. ¶ 148.)

As CDK admits, prior to September 2013, CDK recognized that maintaining an "open" DMS limited the revenue that it could generate if data could instead be

accessed only through CDK's own certified access program, 3PA. (Id. ¶ 163 (citing internal CDK presentation discussing how blocking "unauthorized code" on the CDK DMS could result in "over $14 million in potential annual revenue to the 3PA business").) Around October 2014, CDK began developing a "refreshed" 3PA program under which it planned to raise prices.¹⁵ (Id. ¶ 164.)

15 CDK asserts, via an employee's testimony, that it had not raised 3PA prices since 2000, although CDK did not produce data to verify her statement. (PRDSOF ¶ 164.)

5. Reynolds's restrictions on DMS access leading up to the alleged conspiracy period

In 2006, after a merger with a relatively more "closed" DMS competitor, Reynolds began taking steps to close its own (previously "open") system by interrupting access by independent integrators. (PRDSOF ¶ 80.) As explained in a September 2006 notice to dealers, Reynolds began releasing a "series of software security enhancements" to "reduce the risk of security breaches by securing [the dealer's] system from unknown access" and "decrease the likelihood of unauthorized data access or loss." (Id. ¶ 82.) The notice advised dealers that "Reynolds will continue to take these steps to remove such access." (Id.) The parties dispute how effective Reynolds was in "closing" its system prior to 2013 and whether a "closed" system is necessarily more "secure." (See id. ¶¶ 80-81.)

Contemporaneous internal Reynolds documents assert that the "primary objective" of these measures was "to ensure the dealer's data is secure and monitored" and that Reynolds needed to "redefine how dealer data may be accessed and protected" because there were "too many 3rd party companies accessing ERA servers in an unauthorized manner." (Id. ¶ 84.) According to Reynolds, these types of measures were designed to "[p]rotect dealer data from inappropriate use by third-party vendors"; "[c]omply with state and federal privacy and consumer-data laws"; "[d]ecrease chances for unauthorized data access or loss"; and "[k]eep Reynolds's ERA system performing optimally." (Id.) Plaintiffs maintain that Reynolds's stated reasons do not reflect its true motivation for restricting access—that is, according to Plaintiffs, increasing profits—and dispute that dealer-authorized data integrators were "unauthorized" or posed a security risk. (See id.)

Between 2009 and 2013, Reynolds implemented several technical measures designed to prevent automated access to its DMS; for example, it disabled third-party code on its servers and introduced CAPTCHA prompts. (Id. ¶ 83.) Plaintiffs dispute that these measures in fact prevented independent integrators from accessing Reynolds DMS data on the dealers' behalf—citing, for example, an internal CDK presentation from February 2012 in which CDK claimed that Reynolds "has still not successfully locked out the major data providers (DMI, IL, Superior, Homenet, Authenticom, etc.) after years of trying." (Id. (citing [1083-22] at 13).)

There are at least four circumstances when Reynolds admits it "whitelisted"—that is, exempted—third parties from certain security measures: (1) after a vendor signed up for RCI but before the certified interface was ready; (2) during an agreed wind-down process when data integrators were serving out the remaining terms of their existing vendor contracts; (3) during end-of-life transitions for certain products; and (4) for "certain specific needs for large dealership groups." (Id. ¶¶ 89, 91-92.) Plaintiffs maintain that, in actual practice, Reynolds whitelisted login credentials in many more circumstances. They point to testimony and other records showing that,

before February 2015, Reynolds exempted tens of thousands of data integrator user IDs because of business concerns. (Id. ¶ 89.) They also cite records showing that Reynolds whitelisted some accounts for more than a year and whitelisted "on demand" for its largest and most influential dealership group clients. (See id.; see also DRPSOF ¶ 27.) Defendants admit that, during major dealers' renewal periods, Reynolds exempted some of those accounts from security measures. (DRPSOF ¶ 28.)

Reynolds heightened its blocking efforts beginning in February 2013, when it announced a "new enhancement" to its software that periodically identified and disabled "suspicious User IDs" that it determined were using non-Reynolds-supported methods to access the DMS. (PRDSOF ¶ 85.) That enhancement caused Authenticom user IDs to be disabled "en masse" starting in June 2013. (Id. ¶ 86.) Indeed, Authenticom CEO Steve Cottrell testified that, "[o]ver a three-month period in the summer of 2013, Reynolds disabled 27,000 profiles used by Authenticom at more than 3,600 dealers." (Id.) Internally, Authenticom called this the "Reynolds apocalypse," and Cottrell said it "resulted in an almost complete collapse of Authenticom's integration business for dealers using the Reynolds DMS." (Id.)

6. Reynolds's restrictions on DMI's DMS access and the beginning of Defendants' negotiations

Reynolds's blocking efforts in the summer of 2013 initially targeted Authenticom and SIS, but began to affect DMI, one of the two third-party integrators owned by CDK by August 2013. (DRPSOF ¶ 20.) In August, DMI's "success rate" in connecting to the Reynolds DMS dropped from approximately 80 percent to less than 40 percent, prompting DMI declare "a full [Reynolds] crisis." (PRDSOF ¶ 98.)

The parties dispute the consequences of Reynolds's blocking measures. Plaintiffs assert that Reynolds's efforts to block DMI ceased by late August after CDK responded "across all fronts: including daily war room meetings to restore locked User IDs, identify root cause of the blocking, and press OEMs to apply pressure against Reynolds." (DRPSOF ¶ 32; PRDSOF ¶ 99.) DMI also solicited dealers to pressure Reynolds to stop its blocking efforts, and had some success in this campaign: by late August, Reynolds ended disruptions because of the backlash, according to CDK's Gardner. (See DRPSOF ¶ 24.) Defendants, for their part, point to internal emails indicating that CDK did not believe that OEM and dealer pressure was a viable long-term solution to Reynolds's blocking. (PRDSOF ¶ 99.) As Plaintiffs see things, this dispute about access to Reynolds' DMS set the stage for an agreement between CDK and Reynolds to prevent the companies from dragging one another down.

Plaintiffs cite August 2013 emails and other communications showing that dealers were threatening to stop doing business with Reynolds. (DRPSOF ¶ 33 (Reynolds salespersons reporting that Reynolds was being "buried" with dealer calls and facing "numerous short-term customers threatening to leave.").) In early September, a Reynolds employee named Ken Bunney reported to senior executives, including Brockman, that September would be the "poorest month in a couple years" and that Reynolds was "looking at Q3 being the poorest quarter in a couple years also." (See [1090-48].) On the other hand, Defendants note evidence that by mid-September, Reynolds had survived the storm: as of that time, eight major OEMs had agreed to—or were actively negotiating—certified data agreements with RCI and five others were "in queue" to do so. (See PRDSOF ¶ 101.) Plaintiffs respond that

RCI's progress signing OEMs was possible only because the conspiracy—discussed below—had begun and CDK by that point was "helping [Reynolds] with OEMs" and "promoting that they should move to RCI," according to notes from Howard Gardner of CDK. (Id.)

In August 2013, before the alleged agreement, Gardner warned in an internal email that absent a "'business' solution"—in other words, a "negotiated outcome" between CDK and Reynolds—CDK would be forced to continue fighting Reynolds, which would harm both sides [1069-183]. Gardner wrote that CDK would "go down in fire and flames rather than surrender the fight" and would face "Hard Landing" or "Rough Landing" scenarios in which Reynolds "continues to lock and relock user accounts," while Reynolds would "lose dealer contract renewals" as CDK "attack[ed]" the Reynolds accounts. (DRPSOF ¶ 34.) In that document, Gardner discussed three scenarios that might play out, absent an agreement; in each scenario he described, DMI would continue attempting to provide data integration services for Reynolds dealers. But Gardner testified, as Defendants point out, that in each of the three scenarios, if CDK could not reach agreement with Reynolds CDK ultimately would have had no alternative other than to "exit[] the hostile integration business." (See id. (quoting Gardner's deposition).)

In August and September 2013, high-ranking CDK and Reynolds executives spoke several times, but all of them denied in depositions having entered into any conspiratorial agreement. (See id. ¶ 35; PRDSOF ¶ 194.)

7. The beginning of the alleged conspiracy and lead-up to the 2015 Agreements

On September 27, 2013, Howard Gardner of CDK met with Reynolds's VP Bob Schaefer in CDK's Columbus, Ohio offices. According to Gardner's internal summary, the two discussed four topics: (i) Market Messaging — Data Security: Reynolds "would like to forge a common perspective with ADP [CDK] on the topic of 'Data Security' messaging," "centered on the [August 2013] NADA security guidelines" to "synchronize our security perspectives and speak with similar voices to the industry"; (ii) Reciprocal 3PA/RCI Access: "contracting with each other using our existing 3PA/RCI contracts, processes, integration technologies, etc."; (iii) a DMI "Soft Landing": "DMI supports the transition of the OEM to RCI from DMI data access technology" and offers "a formal 'white list' protected program ... to our existing [vendor] clients"; and (iv) Reynolds's proposed indemnification language,¹⁶ which Gardner said had "stalled all prior discussions" with CDK. (PRDSOF ¶ 102; see also [975-61].)

16 The court has been unable to ascertain from the record exactly what Reynolds was demanding in terms of indemnification. The final agreement between CDK and Reynolds [975-49] contains a broad indemnification provision under which both parties agreed to indemnify the other for liabilities "arising out of in connection" with the Agreement.

Plaintiffs respond that Defendants "omit key details of Gardner's summary," including that Gardner said the two had reached "'conceptual' common ground" on the "key business points" and that they "discussed and agreed upon a general approach" for DMI to keep accessing the Reynolds DMS. (See PRDSOF ¶ 102; see also DRPSOF ¶ 36.) Plaintiffs also emphasize the full context around market messaging. According to Gardner's summary:

Schaefer opened up with the point that [Reynolds] would like to forge a common perspective with ADP on the topic of

"Data Security" messaging. Much of their interest is centered on the NADA security guidelines. Their goal is to synchronize our security perspectives and speak with similar voices to the industry. We didn't discuss details, but he specifically mentioned coordinated messaging leading into NADA next March. I told him I didn't believe this was out of the question, but is obviously subject to agreement on the specifics. My take is that this aspect of an agreement is very important to them.

(PRDSOF ¶ 102.)

On October 31, 2013, Gardner sent Ron Workman—another CDK executive—a draft of his proposed email to Schaefer summarizing their meeting. (Id. ¶ 104.) Workman initially suggested that Gardner "[o]mit the market message language" from the email entirely, but Gardner responded: "I believe we need to leave some kind of market messaging language in the email. That is a key issue for [Reynolds CEO] Brockman, and if we reach an agreement with [Reynolds] to allow reciprocal integration there will have to be some kind of announcement that we all agree on." (Id.)

On November 12, 2013, Gardner sent the summary to Schaefer, describing it as "a summary of the main business points" for "a possible agreement" between the two companies. (Id. ¶ 105.) Gardner wrote that CDK and Reynolds were "conceptually in agreement on the following" three points: (1) reciprocal 3PA/RCI access; (2) the DMI "soft landing"; and (3) "Market Messaging — Data Security." (Id.) Plaintiffs cite several instances in which, following the meeting between Gardener and Schaefer, "Gardner repeatedly directed his team not to criticize Reynolds's closed-system policies." (DRPSOF ¶ 37.) Defendants respond that the documents merely "reflect DMI's efforts to negotiate a mutually agreeable 'soft landing' in the wake of Reynolds's August 2013 security measures," which strained DMI's business. (Id.)

After the September 2013 meeting, Reynolds significantly reduced the frequency of its efforts to block DMI's use of credentials that enabled it to access the Reynolds DMS. (Id. ¶ 38.) Indeed, while more than fifty percent of DMI user credentials were locked as of October 1, fewer than twenty percent were locked three weeks later. (See id.) Defendants add context—that Reynolds provided some temporary exemptions before September 2013 and exempted other users as well—but admit the general trend.

Plaintiffs point to evidence that, in 2014, CDK coordinated with Reynolds on a marketing strategy to attack data integrators—specifically Authenticom—using "security" as the rationale. (See ¶ 40.) Defendants respond that "CDK began to warn dealers not to share credentials with third parties well before 2014" and did so, Defendants insist, without coordination between CDK and Reynolds. (Id.)¹⁷

17 In February 2014, Malcolm Thorne—who was later selected to serve at the Chief Global Strategy Office of CDK after its spinoff from ADP—proposed ideas for the "new" CDK, including a "[r]e-evaluation of U.S. 3PA business model" involving "[m]ore restrictive access, higher rates/minimums, [and] better client data protection program for our clients' data[.]" (PRDSOF ¶ 151.) Thorne did not directly negotiate the February 2015 contract with Reynolds, but Plaintiffs point to a PowerPoint presentation and testimony from Thorne indicating that he was involved in and kept apprised of the overall strategy and objectives for the parties' agreements. (Id. ¶ 152.)

In any event, negotiations between Reynolds and CDK continued in 2014. In an internal email on April 1 of that year, Gardner summarized another conversation he had with Schaefer. In that conversation, Gardner reported, he had learned that Reynolds intended to "launch new 'security

enhancements'" in the very near future, which would "significantly impact" independent integration on the Reynolds DMS. (PRDSOF ¶ 107.) Reynolds was willing, however, to "cancel the disruption" if CDK would sign an indemnification agreement (on which "little progress" had been made) and accelerate negotiations with Reynolds. (Id.) Gardner had warned Schaefer that "it would be extremely poor form to proceed with the disruption and would set back negotiations." (Id.) Plaintiffs do not dispute that these statements were made, but citing testimony from CDK's economic expert Dr. Whinston, they characterize Schaefer's conduct as "hardballing in the negotiations" to get better terms for Reynolds. (Id.) Plaintiffs also cite Gardner's comment in an internal email that he would "be surprised" if Reynolds "actually moves forward with the disruption" given Schaefer's "cooperative and conciliatory tone." (Id.) Although CDK did not sign the agreement in April, Plaintiffs note that Reynolds in fact never blocked DMI, and its internal records did not reflect a plan to do so. (Id.) That evidence, according to Plaintiffs, aligns with Plaintiffs' belief position that CDK and Reynolds had agreed in September 2013 to work together and that CDK knew Reynolds "planned to release [security] updates to block independent integrators" but only "after the written agreements [between CDK and Reynolds] were signed." (See Id.)

On April 8, 2014, Reynolds sent back a redlined version of a draft interim integration agreement; Gardner told Schaefer that CDK would "sign the agreement 'today' with all their latest changes," except that CDK was "firm" about its demand for "a 30 day notice period" (if Reynolds opted to extend the period for DMI to wind down its access to the Reynolds DMS, presumably meaning that DMI would need to be prepared to continue offering integration services). (Id. ¶ 108; [975-62].) Internally, Gardner said that the notice provision could be an issue and that, without a written agreement, "[w]e'll just fight it out in the DMS world for another two years or more." (PRDSOF ¶ 108.)

As negotiations continued into June 2014, Reynolds's CEO Brockman demanded free access to CDK's systems for 20 years because CDK's predecessor, ADP, "has been extracting data out of Reynolds systems for over a decade"; "ADP has wrongly taken advantage of Reynolds in the marketplace over the issue of data security — that has cost us in the millions"; and CDK "caused a lot of grief with OEMs." (Id. ¶ 109.) CDK refused; in July, Ron Workman from CDK emailed Gardner—and inadvertently copied Schaefer, it seems—and said that giving Reynolds "unfettered access to our system" would not work and that CDK had no choice but to "draw a line in the sand." (Id. ¶ 110.)

Reynolds continued to push the issue: In early July, Schaefer reiterated Reynolds's demand for unfettered access to CDK's DMS, though it reduced the demand from twenty years of access to just five. (Id. ¶ 111.) In response, Workman from CDK declared that the "deal is dead," and CDK's CEO Anenen advised his subordinates that "we have to prepare for [Reynolds] to block our access and work through it." (Id.) Gardner noted that Schaefer's response was a "significant back-track from recent statements he's made to us about the integration they require" and "[e]ither Schaefer isn't in sync with Brockman (and never was), or Brockman has changed the rules." (Id.) Gardner said that DMI must "prepare for a new round of [Reynolds] disruption" because it was "hard to imagine reaching any kind of workable agreement with people who can't be trusted." (Id.) Nonetheless, after Reynolds proposed a five-year "No Fee" term under which its applications would use "the processes [CDK] currently provide[s]

to other 3rd party vendors," negotiations resumed. (Id. ¶ 112.)

According to CDK executives, around that same time, CDK decided to close its system to data integrators. Plaintiffs claim that CDK informed Reynolds about this decision before announcing it publicly. In support, Plaintiffs cite an August 2014 email from Schaefer to other Reynolds executives, in which Schaefer warned that CDK was "getting ready to do a me too" on "security" and suggested that it was "time to promote" Reynolds's security advantage before then [1089-49]. Plaintiffs also cite Brockman's notes from a July 2014 sales meeting, where he wrote about CDK's plans with respect to "their version of RCI," i.e., 3PA. Plaintiffs urge that both of these documents pre-date CDK's public announcement of its plan to close its DMS and require vendors to use 3PA. (DRPSOF ¶ 49.) Defendants argue that these documents reflect speculation and are insufficient to show that CDK told Reynolds of its decision. (See id.)

On September 8, 2014, CDK's Gardner summarized the proposed agreement with Reynolds in a draft presentation for CDK senior executives, which contained the following slide:

ADP - R&R Integration Agreement

Strategic Rationale

R&R Strategic Goals

1. Terminate their agreement with Authenticom to provide hostile integration to ADP dealers.
2. Massively disrupt Authenticom: ending viability as a hostile integrator to R&R DMS.
3. Agreement with ADP is path to fully securing R&R DMS with least dealer/market disruption.
4. Maintain continuity of service for R&R applications/services installed in ADP dealerships.
5. Can show their actions are in harmony with their 'Certified Data' messaging.

ADP Strategic Benefits

1. RCI certification for ADP applications enables new growth opportunities in R&R segment.
2. Relieves pressure on current OEM programs (Cobalt).
3. Opens up opportunities to participate in new OEM sponsored programs (e.g. Lot Management)
4. Eliminates uncontrolled revenue losses and brand destruction at DMI as R&R secures their DMS.
5. Opens new opportunities to increase DMI revenue as R&R disrupts other hostile integrators.
6. Puts greater pressure on Authenticom/DealerVault.
7. Marketing "halo" of solving the data access problem in the industry.
8. Can demonstrate that our actions are consistent with our 'Approved Access' messaging.

(Id. ¶ 53.) The next day, after another CDK executive had "some feedback" to offer Gardner "that may be best suited for a call," Gardner distributed a revised version of the slide deck to a larger group within CDK. (See id.) That later presentation eliminated the reference to "put[ting] greater pressure on Authenticom/DealerVault" and edited the reference to "massively disrupt[ing] Authenticom," changing it to read: "Flexibility to increase disruption of hostile integrators." (Id.)

In October 2014, after the spinoff, Gardner instructed CDK employees to begin working on "a significantly revised CDK 3PA Strategy," including a "Secure DMS

Strategy." (PRDSOF ¶ 155.) That strategy, outlined in March 2015, identified five reasons for securing the DMS: (1) "Protect CDK," (2) "Protect our customers," (3) "Improve the CDK Client Service Experience," (4) "Reduce Product Support Costs," and (5) "Execute on our Business Strategy." (Id. ¶ 156.) This outline did not reflect the real reason CDK "closed" its system, according to Plaintiffs. They cite internal records suggesting that CDK's goal in closing its DMS was to raise 3PA prices, and that CDK felt threatened by "disintermediation"—that is, the threat that third-party software applications might take over functions traditionally performed by the DMS. (See id. (citing, among other things, DRPSOF ¶¶ 84, 120).) Plaintiffs also point to testimony and internal CDK documents suggesting CDK had decided to close access to its DMS by summer 2014. (See DRPSOF ¶ 48.)

By late 2014, CDK and Reynolds were negotiating "contract structures and details." (PRDSOF ¶ 112.) As late as February 16, 2015, the parties were still debating agreement terms, minimum monthly payments, and phase-in pricing terms. (Id. ¶ 115.) In an email outlining these issues, CDK's Workman told Schaefer at Reynolds that "if you want to get these agreements completed then let's get reasonable with the final issues[.]" (Id.)

8. The February 2015 Agreements

On February 18, 2015, CDK and Reynolds signed the 2015 Agreements: three written agreements—the Data Exchange Agreement ("DEA"), the 3PA Agreement, and the RCI Agreement—described in more detail below. (PRDSOF ¶ 116.)

a. Data Exchange Agreement

Under the DEA, DMI could continue extracting data from the Reynolds DMS "without interruption from Reynolds security enhancements" during a "Wind-down period," which lasted until the later of: (1) the end of each DMI client's current term with DMI "up to a maximum of one year"; or (2) May 13, 2015. (PRDSOF ¶ 119.) Reynolds could extend the "wind-down period," at its discretion, for up to five years from the DEA's effective date. (Id.) During the wind-down period, CDK agreed to "reasonably cooperate with Reynolds's efforts" to have DMI's vendor clients "execute agreements to become part of the Reynolds RCI Program." (Id. ¶ 120.)

CDK agreed to inform DMI clients that it intended to "wind down" its "Reynolds DMS related integration" and did not intend to "extend service for such integration beyond the current primary term of its agreement." (Id. ¶ 121.) Plaintiffs point out that CDK also agreed: (1) to "express[] words to the effect that CDK looks forward to doing its part to ease each CDK Third Party Client's transition away from CDK's integration with respect to dealers using the Reynolds DMS"; and (2) to provide Reynolds with "contact information" for CDK's vendor clients "for purposes of any inquiries regarding" RCI. (Id.) In March 2015, CDK sent letters to its DMI vendor clients—after obtaining approval and comments from Reynolds—stating that the "RCI program has proven to be a successful approach to consistent, reliable exchange of data from a [Reynolds] DMS." (Id.; see also DRPSOF ¶ 57)

CDK agreed, in the DEA, to provide a list to Reynolds of the DMI user logins that would need whitelisted access to Reynolds DMSs. (PRDSOF ¶ 122.) Plaintiffs emphasize that the agreement also provided that CDK was entitled to "continue providing its integration services through DMI in accordance with its contracts without interruption from Reynolds security enhancements" and that Reynolds agreed not to "take measures to block or otherwise disrupt DMI's normal Reynolds DMS access." (Id.) CDK and Reynolds agreed not to "sell, transfer or assign" any "technology, business process, or other such knowledge" about how to integrate with the other's DMS. (Id. ¶ 123.) CDK and Reynolds also agreed each would take no other steps "to assist any person that it reasonably believes to have plans to access or integrate with the other party's DMS without the other party's written consent." (Id.) That section of the agreement included the assertion that it was "not intended as a 'covenant not to compete,' but rather as a contractual restriction of access and attempted access intended to protect the operational and data security integrity" of the Reynolds and CDK DMSs. (Id.) That "contractual restriction of access" never expired. (See id.) Reynolds discounts the significance of this language; Schaefer testified that Reynolds never in fact agreed to a "contractual restriction of access and attempted access" and said the provision "shouldn't have been in there" because "there was no agreement for CDK not to come into [Reynolds's] system." (DRPSOF ¶ 121.) Schaefer blamed the attorneys who drafted the agreement; he also claimed that the title of the Section—"Prohibition on DMS Access"—should have been deleted. (Id.) Schaefer has not said that he failed to read the agreement before it was executed; he merely noted for the first time in his deposition that these were "mistakes."

b. The 3PA and RCI Agreements

Under the 3PA Agreement, also executed on February 18, 2015, Reynolds agreed that it would no longer access the CDK DMS through any "third party," including via any "hostile interface," except that Reynolds could continue to receive data from dealers if that data were provided to Reynolds "without the use of any third party service or application." (DRPSOF ¶ 56; see also [975-50].) CDK also agreed that for five years—and for up to 600 CDK dealership rooftops¹⁸—Reynolds applications would have access to dealers' data without payment of the standard 3PA fees (the "No Fee" term). (PRDSOF ¶ 129.) After that "No Fee" term expired—and for any rooftops exceeding the 600-dealership cap—Reynolds agreed to pay CDK's standard 3PA fees. (Id.) CDK made a reciprocal commitment: CDK would no longer use unauthorized methods to access the Reynolds DMS, but could still obtain data directly from Reynolds dealers without use of a third-party service. (DRPSOF ¶ 56.) Pursuant to the RCI Agreement, a list of CDK's proprietary applications became eligible for the RCI program. (PRDSOF ¶ 131.) In other words, Reynolds agreed to provide CDK's applications with certified access to Reynolds' DMSs (via RCI) for a fee.

18 A "rooftop" in industry lingo means "each physical dealership location." (See Expert Rep. of Michael D. Whinston [975-78] ¶ 42.) A "dealer" may have a single rooftop—i.e., only one location—or a set of rooftops under common ownership. Id.

9. Defendants' security efforts after February 2015

a. CDK

Five days after the 2015 Agreements were executed, Malcolm Thorne—CDK's Chief Global Strategy Office—instructed CDK's research and development department to begin developing methods to "secure the platform ... GET THIS GOING IMMEDIATELY." (DRPSOF ¶ 60.) Securing its DMS became part of CDK's broader strategy known as "SecurityFirst," which it publicly announced in June 2015. (PRDSOF ¶¶ 157-58.) CDK maintains—based on testimony from Brockman, Schaefer, and another Reynolds employee in the company's data services department—that CDK did not discuss its plans with nor otherwise give

advance notice to Reynolds prior to that announcement. (Id. ¶ 158.) Plaintiffs dispute this (see id.), citing among other evidence Brockman's notes from a July 2014 sales meeting that CDK planned to implement "their version of RCI." (DRPSOF ¶ 49.)

CDK presented aspects of the SecurityFirst initiative to a group of dealership CFOs in April 2015. CDK's Chief Marketing Officer, who presented to the group, reported internally that "the program was well received," but Plaintiffs dispute that dealerships supported SecurityFirst. (PRDSOF ¶ 157.) Plaintiffs cite evidence that SecurityFirst was considered a marketing campaign designed to drive up 3PA revenue, rather than a bona fide security effort. (See id.) For instance, an internal email from CDK's Global Security Officer, Josh Sowers, criticized the company for launching its initiative "where our security sucks and is no where near industry best practices." (DRPSOF ¶ 122.) In another email, Sowers complained that CDK had "[n]o product security strategy and Yet we want to talk security to sell products." (Id.) Sowers believed CDK was "going to market way too soon" and could not "answer the fundamental question of what is more secure." (Id.) Plaintiffs also note skepticism on the part of some vendors and dealers about whether the new measures were genuinely adopted for security reasons. (Id. ¶ 125; see also id. ¶ 126 (evidence from 2006 to 2013 that CDK employees believed Reynolds's similar security measures were motivated by financial rather than security concerns); id. ¶ 128 (evidence CDK relaxed security measures for dealers whose DMS contracts would be up for renegotiation soon).)

CDK did not raise prices for 3PA vendors until the second half of 2015. (PRDSOF ¶ 165.) In March 2016, CDK first began using technological measures to prevent access to its DMS by independent integrators. (Id. ¶ 159.) CDK implemented CAPTCHA controls for certain DMS accounts suspected of unauthorized third-party access in October 2017. (Id. ¶ 161.) Authenticom attempted a work-around, using computer scripts that could in theory re-establish access to CDK's DMS despite those technological measures. But according to Cottrell (Authenticom's CEO), the scripts were largely unsuccessful, and Authenticom today provides automated data integration services for virtually no CDK dealerships. (Id. ¶ 162.)

b. Reynolds

On February 20, 2015, representatives of CDK and Reynolds met to discuss methods for whitelisting (i.e., granting access permissions to) DMI credentials during the Wind-down period, which was essential to DMI continuing to provide data integration services on Reynolds' DMSs. (PRDSOF ¶¶ 124-25.) A few weeks later, around March 16, Reynolds released more security enhancements that would prevent unauthorized automated access by Authenticom and other third parties. (DRPSOF ¶ 58.) Shortly thereafter, Reynolds sent Authenticom a cease-and-desist letter demanding that Authenticom stop accessing the Reynolds DMS without authorization. (PRDSOF ¶ 135.) Reynolds offered a Wind Down Agreement to Authenticom in May, which Authenticom declined. (Id.)

Other integrators did wind down their business on the Reynolds DMS. For example, SelectQu—owned by Dominion, another DMS proprietor—abandoned efforts to provide third-party integration sometime in 2015 or 2016. This decision was based, in part, on the fact that Reynolds's security measures had become too onerous to consistently overcome but also (according to Dominion's corporate representative) a response to Reynolds and CDK's demand that Dominion abandon its independent integration business before they would certify

Dominion's layered (i.e., proprietary) apps for access to RCI and 3PA. (Id. ¶ 134.)

10. Cottrell's conversation with Defendants' executives

Authenticom's Cottrell testified that in May 2015 he spoke by telephone with Schaefer at Reynolds. (PRDSOF ¶ 185.) In that conversation, according to Cottrell, Schaefer "restated that Mr. Brockman was adamant about removing all third parties from the Reynolds system. He — his job and what he was doing was forming agreements and alliances with all of the major DMS players and the OEMs, okay, to only use certified interfaces provided from the DMS providers and to support each others' activities in that direction." (Id.) Cottrell testified that by "the major DMS players" he "certainly understood it to mean CDK." (Id.) According to Cottrell, Schaefer's message was: "Resistance is futile, surrender, get out of the business." (Id.)

About a year later, in April 2016, Cottrell spoke to CDK's Vice President of Product Management, Daniel McCray, at the NADA convention in Las Vegas. (Id. ¶ 188; see also DRPSOF ¶ 62.) The following day, Cottrell made notes of that conversation, making a record of everything said by McCray that Cottrell deemed important. (PRDSOF ¶ 188.) The notes reflect twelve "Statements from Dan":

• "We have clocked you in over 2000 of our stores CDK will not allow you to continue to exploit our systems for profit."

• "You are taking away our revenue stream."

• "You should take our deal [while] you still have something left to sell."

• "Your book of [Reynolds] business is worthless to us because [] the agreement between CDK and Reynolds makes that business worthless to us."

• Most of the major DMS companies have agreed to only source data from each other and effectively "Lock you and the other third parties out."

• "I wanted to look you in the eye and let you know man to man, I have been mandated by our new CEO to seek you out and destroy your business on our systems."

• "We removed the code from Edmonds on our boxes last week."

• "Our contract with our dealers does not allow them to give you a user name and password to continue to export data."

• "We will enforce our contract with dealers and sue them if needed to keep you out of our systems."

• "We have had over 50 of your customers in our booth in the last few days and anticipate signing contracts with many of them."

• "I have less than 180 days before I retire, and I will get this accomplished before I go. Whatever it takes."

• "For god's sake, you have built a great little business, get something for it before it is destroyed otherwise I will F'ing destroy it."

(Id. ¶ 189; see also DRPSOF ¶ 62.)

Authenticom COO Brian Clements testified that Cottrell told him about this conversation with McCray at the NADA convention shortly thereafter. (DRPSOF ¶ 64.) Other attendees of the NADA convention likewise reported stories of McCray "yelling at folks." (Id. ¶ 65.) McCray admits that he spoke with Cottrell at the convention but denied referencing any CDK-Reynolds agreement or making a profane threat to destroy Authenticom's business. (Id. ¶ 66.) CDK was considering buying part of Authenticom's business around the time of the McCray-Cottrell

conversation, though it ultimately opted not to. (PRDSOF ¶ 191.)

11. Competitive effects of the alleged conspiracy

a. DMS switching

It is undisputed that CDK and Reynolds competed to some degree to sell DMS software during the alleged conspiracy. (See PRDSOF ¶ 174.) To win business, some DMS providers—including CDK and Reynolds—have discounted or subsidized part of the out-of-pocket expenses associated with switching DMS, but Plaintiffs dispute that subsidies cover "anything close to the full costs of switching," meaning that there are still strong cost disincentives for a dealer to switch. (Id. ¶ 175.)

In support of their position that the alleged conspiracy made the DMS market less competitive, Plaintiffs point to the rates at which DMS customers "switched" from Reynolds to CDK. Plaintiffs believe that Defendants' anticompetitive agreement pushed switching rates even lower than in a hypothetical no-collusion world. Defendants reject that position. Looking to data from 2008 to 2018, Defendants note that switching peaked in 2011 and declined in all subsequent years, except for a slight increase from 2013 to 2014 and from 2017 to 2018. (Id. ¶ 172.) Although Plaintiffs do not specifically object to that summary, they argue it presents an incomplete and inaccurate view of the data. According to AutoLoop's expert (Dr. Israel), in August 2013, Reynolds initiated large-scale lockouts of independent data integrators and experienced the worst month for its DMS business in years. (Id.) Despite these blocking efforts, Reynolds's losses to CDK did not increase, which, according to Israel, "confirms that Reynolds succeeded in cooperating with CDK to prevent a spike in losses to CDK following its 2013 lock-out initiatives." (Id.) After 2014—when CDK began to target independent integrators and took steps to close its own system—Reynolds's losses to CDK fell still lower. (Id.; see also id. ¶ 170.)

Between September 2013 and December 2018, Reynolds won hundreds of rooftops from CDK, and CDK won hundreds of DMS rooftops from Reynolds. (Id. ¶ 173.) But those numbers should not be interpreted as establishing robust competition, Plaintiffs contend, because Reynolds was losing hundreds of customers per year to CDK in 2008-2012 and lost 52 or fewer customers to CDK in 2015, 2016, and 2017. (Id.; see also DRPSOF ¶ 39.)

b. RCI and 3PA prices

CDK management recognized that preventing independent access could increase the 3PA's profitability. (DRPSOF ¶ 120.) Plaintiffs—relying in part on reports from their experts, Drs. Israel and Williams—claim that after excluding competition from independent integrators, Reynolds and CDK raised their certified integration prices substantially above competitive levels. (Id. ¶ 72.) Defendants deny that they excluded competition, but admit that in the second half of 2015, CDK increased its 3PA prices for the first time since 2000. (Id.) Whether CDK offered any quality improvements in return for the increased prices is disputed. (See id. ¶ 75.) Reynolds also increased prices for RCI in September 2013, but announced those increases in June. (See id. ¶ 73.)

Other DMS providers charge between $25 and $75 a month for data integration services; the parties dispute whether those services are as good or better than 3PA and RCI, but neither side supplies definitive evidence on this issue. (See id. ¶ 78.) Plaintiffs maintain that no DMS providers other than CDK and Reynolds block dealers from using data integrators; Defendants respond that at least two small DMS companies—Talon and NowCom—have, at times, also prohibited third-party data integration. (See id. ¶ 79.) c. The app market

Plaintiffs assert that Defendants suppressed competition in the app market in several ways. First, Plaintiffs contend, by blocking data integrators and raising data integration charges to supracompetitive levels, CDK and Reynolds drove out providers of apps (and prevented entry from new apps providers) that could not afford the inflated 3PA and RCI fees. Plaintiffs cite a few specific examples as well as generalized testimony concerning the state of the market. (See DRPSOF ¶ 80; see also id. ¶ 83 (Plaintiffs' assertion that inflated 3PA and RCI fees deprive vendors of resources for innovation and expansion, supported by testimony from two vendors' corporate representatives).) Defendants dispute that their fees were supracompetitive or that they drove out new apps; they cite as an example the testimony of AutoLoop's Vice President that AutoLoop has "more competitors" in 2018 than in 2015. (Id. ¶ 80.) Defendants also cite the experience of Plaintiff MVSC, which has captured 60 percent market share without signing up for RCI or 3PA. (Id.)

A second way that Defendants suppressed competitions, Plaintiffs claim, is by giving superior service and functionality to Defendants' own apps. For example, they cite the undisputed fact that, as part of its 3PA "refresh" and SecurityFirst initiatives, CDK rendered some third-party applications unable to open repair orders in the DMS. (Id. ¶ 81.) Defendants dispute that vendors actually lost business because of that change. (Id.) Plaintiffs also point to testimony from vendors of third-party apps that they lost some data functionality when they switched from independent integrators to 3PA and RCI. (See id. ¶ 82.)

Both CDK and Reynolds decline to do business with vendors they deem too small to participate in the 3PA and RCI programs. Plaintiffs contend this too, undermines competition by depriving small and start-up vendors of access to automated data integration services. (Id. ¶ 85.) Defendants respond that during the relevant period, CDK offered reduced monthly fees in a special program it offered to vendor startups. (See id.) Plaintiffs nevertheless assert that Defendants' access restrictions have harmed innovation, citing a presentation by NADA's Brad Miller at the 2018 NADA conference:

[T]he biggest thing I hear from dealers [is] a frustration that they can't provide the customer experience that they want because they can't get the information back and forth, right? ... [F]olks have to enter data nine and 10 times in the course of selling a car. And that's crazy. It increases customer frustration. It paints dealers with this brush that we're technological dinosaurs.

....

[T]he way that we run customers through the service line hasn't changed in 25, 30 years and so there's an increasing frustration. This [is] all tied to the information you have and how it gets back and forth with third parties, how it is shared, how it is stored, how do systems talk to each other.

(Id. ¶ 86.)

d. Authenticom

Citing statements and deposition testimony from employees, vendors, and CDK's McCray, Plaintiffs maintain that Defendants aimed to drive Authenticom out of the market starting in 2013. (DRPSOF ¶ 87.) Defendants dispute this; they submit that CDK did not block Authenticom until March 2016 and that Authenticom still does significant business on the Reynolds platform. (Id.)

Authenticom's expert, Ms. Lawton, opined that Authenticom's growth on both the CDK and Reynolds systems slowed after Defendants' alleged September 2013

agreement. (Id. ¶ 89.) After the 2015 Agreements, Authenticom's connections on both systems decreased sharply again, and Authenticom's CDK connections decreased even more sharply in mid- to late-2016 after CDK began blocking Authenticom's access. (See id.)

According to Dr. Israel's own analysis, Authenticom's CDK connections increased from approximately 6,000 in September 2013 to roughly 9,000 in August 2015, and they did not fall below September 2013 levels until November 2016. (PRDSOF ¶ 36.) And Authenticom's connections to Reynolds DMSs have consistently increased since early 2016 and are at an all-time high today. (See id.) As Plaintiff AutoLoop notes, however, Lawton testified that Authenticom is "worse off as it relates to [Reynolds] because the trajectory that it was on prior to September of 2013 was interrupted.... [T]here are fewer connections than Authenticom would have had in the but-for world as it pertains to [Reynolds] as well as CDK." (Id. ¶ 37.) Similarly, Authenticom receives more revenue today from its business on the Reynolds DMS than at any prior time, but Lawton opined that Authenticom's Reynolds-based revenue "is lower than what it would have been but for the liability contentions in this case." (Id. ¶ 38).

Since at least early 2016, the "vast majority" of Authenticom's business on the Reynolds DMS has been through "dealer-pushed data"—i.e., DMS data that dealers extract using Reynolds's manual reporting tools and then send to Authenticom—rather than through automated access. (Id. ¶ 39.) The parties proffer conflicting testimony concerning whether dealer-push solutions are "painless" or an adequate substitute for automated integration. (See id. ¶ 40; see also DRPSOF ¶ 30.) Taking a few examples, Gardner at CDK said manual-push solutions are "complex and time consuming to set up, they place a high ongoing burden on the dealer, and they are inherently unreliable." (PRDSOF ¶ 40.) CDK's Senior Director for Sales Engineering testified that "manual push is not a viable alternative for data integration, period." (Id.) And Reynolds executive Ed Thornhill described the Reynolds manual tool as "really only a minimally viable solution because it requires manual effort by the dealer each day to create and move the files." (Id.)

According to Lawton, Authenticom charges the same price for dealer-push-driven service as it historically charged for automated polling. (Id. ¶ 41.) Put another way, Authenticom charges the same rate—typically, $50 a month per dealership—for all clients. (Id.) As of August 2013, Authenticom's annualized, twelve-month net income was $3.1 million and as of April 2019, that number was $3.8 million. (Id. ¶ 42.) But while Authenticom's income did grow slightly over that period, Lawton contends its growth trajectory rapidly changed after the alleged conspiracy began. (DRPSOF ¶ 90.) After seeing consistent and rapid growth from its founding in 2002 until 2013, "Authenticom['s] financial performance experienced a marked decline," with annual revenue falling from $17.2 million in 2015 to $11.6 million in 2018. (Id.) In February 2018, Authenticom laid off more than 60 percent of its staff. (Id.) Lawton calculates that Authenticom has suffered at least $115.8 million in lost profits because of Defendants' conduct. (Id.)

e. Defendants' market shares

In addition to CDK and Reynolds, other DMS providers, including DealerTrack, Auto/Mate, Autosoft, DealerBuilt, and Dominion Dealer Solutions, compete to provide services for franchised automobile dealers, though Plaintiffs claim these DMS providers cannot compete for the largest dealership groups—i.e., those with ten or more rooftops. (PRDSOF ¶ 166.) Reynolds emphasizes that its own market share has fallen, a decline that both pre-dates and post-dates the alleged conspiracy. (Id. ¶ 167.) In 2013, CDK's and Reynolds's shares of DMS business with franchised automobile dealers in the United States were approximately 50 percent and 31 percent, respectively. (Id.) Between 2008 and 2018, Reynolds's share of the DMS market decreased by at least 15 percentage points in terms of "rooftops." (Id.) Still, as Plaintiffs observe, there is evidence that Defendants' market share is much higher when measured by sales volume or percentage of the largest dealership groups. (See id.)

Defendants have offered a chart that they believe sets forth the approximate share of franchised dealer rooftops in the United States for CDK, Reynolds, and other DMS providers from 2008 to 2019. This chart, they contend, undermines any suggestion that they engaged in anticompetitive activity in 2013 and thereafter:

Year	CDK	Reynolds	Cox-Dealertrack	Other
2008	-	42%	-	-
2009	48%	40%	-	-
2010	49%	37%	-	-
2011	51%	34%	-	-
2012	49%	32%	-	-
2013	50%	31%	-	-
2014	49%	29%	-	-
2015	49%	28%	13%	11%
2016	48%	28%	14%	10%
2017	47%	27%	15%	10%
2018	47%	27%	18%	9%
2019	47%	25%	19%	9%

(Id. ¶ 168.) Plaintiffs dispute this, as well; they contend the chart shows Reynolds's market share for just one of its two DMS products. In 2017, for example, the chart lists Reynolds's market share as 27 percent, but Plaintiffs claim Reynolds's market share was 30 percent for its two DMS products combined. (Id.) Plaintiffs also repeat their point that Reynolds has higher market share when measured by revenue or among the largest dealership groups. (Id.)

Moreover, according to Plaintiffs, even if Reynolds has lost market share over time, the evidence shows that Reynolds dealers switched away from Reynolds most frequently in the 2011-2012 period (when Reynolds faced competition on openness from CDK) but less frequently after September 2013, when CDK allegedly conspired with Reynolds to stop competing on openness. Plaintiffs provide the following figure from their expert Dr. Bresnahan to illustrate:

Figure 4

Estimated Competitive Loss Rate Among Reynolds Customers With Contracts Up For Renewal

2008-2018

Image materials not available for display.

(Id. ¶ 170.) Plaintiffs also maintain that Reynolds lost a substantial number of rooftops from 2008 to 2014 due to its "closed" data access policies, suffered smaller losses in 2015, and even gained rooftops in 2016, as illustrated in this chart by Defendants' expert (Whinston) (id. ¶ 170):

Figure 38: Net US franchised new rooftop wins by Reynolds

Image materials not available for display. B. Sherman Act § 1 Conspiracy

Section 1 of the Sherman Act "is designed to prevent businesses from entering into collusive agreements." Omnicare, Inc. v. UnitedHealth Grp., Inc., 629 F.3d 697, 705 (7th Cir. 2011). It prohibits "[e]very contract, combination ... or conspiracy, in restraint of trade or commerce," 15 U.S.C. § 1, "though courts have long restricted its reach to agreements that unreasonably restrain trade." Omnicare, 629 F.3d at 705 (citing State Oil Co. v. Khan, 522 U.S. 3, 10, 118 S.Ct. 275, 139 L.Ed.2d 199 (1997)). To prevail under § 1, plaintiffs "generally must prove three things: (1) that defendants had a contract, combination, or conspiracy ('an agreement'); (2) that as a result, trade in the relevant market was unreasonably restrained; and (3) that they were injured." Id.

For the reasons explained below, AutoLoop's § 1 claim survives this motion, but the court grants judgment on the Dealers' claim for damages. Though they may be entitled to injunctive relief, their damages claims are barred by the Illinois Brick doctrine.

1. Element 1: Agreement

Plaintiffs in a § 1 case must "find and produce evidence that reveals coordination or agreement." Kleen Prods. LLC v. Ga.-Pac. LLC, 910 F.3d 927, 936 (7th Cir. 2018). To show "concerted action," a reasonable jury must be able "to infer that the alleged conspirators had a conscious commitment to a common scheme designed to achieve an unlawful objective." Omnicare, 629 F.3d at 706 (quotation omitted). The "less plausible the charge of collusive conduct," the "more evidence is required." In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 661 (7th Cir. 2002). Although on summary judgment the court draws favorable inferences for the non-moving party, "antitrust law limits the range of permissible inferences from ambiguous evidence in a § 1 case." Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986). Thus, "conduct as consistent with permissible competition as with illegal conspiracy" cannot "support an inference of antitrust conspiracy." Id. For that reason, to survive summary judgment, the plaintiff's evidence must "tend[] to exclude the possibility that the alleged conspirators acted independently." Id. (quotation omitted).

This does not mean, however, that the plaintiff has "to exclude all possibility" that the defendants acted independently—that "would amount to an absurd and legally unfounded burden to prove with 100% certainty that an antitrust violation occurred." Toys "R" Us, Inc. v. F.T.C., 221 F.3d 928, 934-35 (7th Cir. 2000) (quotation omitted). There must simply "be some evidence which, if believed, would support a finding of concerted behavior." Id. at 935. The evidence should be examined holistically; it is a "trap" to "suppose that if no single item of evidence presented by the plaintiff points unequivocally to conspiracy, the evidence as a whole cannot defeat summary judgment." High Fructose Corn Syrup, 295 F.3d at 655.

Like all plaintiffs, those bringing antitrust suits can meet their burden "with either direct or circumstantial evidence." Kleen Prods., 910 F.3d at 934. Demonstrating an unlawful arrangement is, of course, "much easier if there [is] a smoking gun," like an admission by an insider. Omnicare, 629 F.3d at 706. More typically, however, evidence is inferential—"economic evidence suggesting that the defendants were not in fact competing" and "noneconomic evidence suggesting that they were not competing because they had agreed not to compete." High Fructose Corn Syrup, 295 F.3d at 655. "While no single piece of information may win the day, the whole may be greater than the sum of its parts in

tending to exclude the possibility of conscious parallelism." Kleen Products, 910 F.3d at 934.

The parties debate (at some length) whether Plaintiffs have presented direct or merely circumstantial evidence. (See, e.g., Defs.' Authenticom MSJ [966] at 43-50; Authenticom Opp. [1071] at 29-31; Defs.' Authenticom Reply [1120] at 3-6.) But the Seventh Circuit has called the distinction between direct and circumstantial evidence "largely if not entirely superfluous"—the ultimate question is whether the evidence "considered as a whole and in combination with the economic evidence" is sufficient. High Fructose Corn Syrup, 295 F.3d at 661-62. That said, strong direct evidence makes it more likely the defendants were not acting independently. Cf. Areeda & Hovenkamp ¶ 308i ("Though factual doubts arise most often when the evidence is circumstantial, the Matsushita principle holds that there is no genuine issue unless the whole record, including direct and circumstantial evidence, makes a necessary fact more likely than not in the mind of a reasonable juror."). For now, the court can sidestep the debate about whether Plaintiffs have offered "direct" evidence. However one characterizes the conversations, documents, and testimony in the summary judgment record, that evidence considered in the aggregate is enough to show agreement.

CDK primarily argues that Plaintiffs' evidence does not tend to exclude the possibility that CDK and Reynolds independently decided to secure their systems against independent access. As CDK frames the argument, Plaintiffs' "claims hinge on whether [they] can present evidence that CDK restricted third-party access to its DMS only pursuant to an agreement with Reynolds." (Defs.' Authenticom MSJ at 51.) In CDK's view, that position is "contradicted by the extensive evidence that Reynolds and CDK were not coordinating on data access policies or any other horizontal elements of competition" but instead were "engaged in independent efforts to defend their interests." (Id. at 51-52 (internal quotation marks and citations omitted).) That argument may well be persuasive to a jury; but at this stage the court is required to view the record and resolve material factual disputes in Plaintiffs' favor. Viewed through that lens, the evidence satisfies Plaintiffs' burden. The court first addresses the most illuminating evidence and then considers CDK's remaining arguments.

a. Evidence tending to rule out the possibility that Defendants acted independently

For years prior to the start of the alleged conspiracy, Reynolds and CDK had competed vigorously over "openness." CDK reminded dealers that they owned their data and should be able to access it and give access to whomever they wanted. CDK used openness to differentiate itself (and take customers) from Reynolds.¹⁹ At his deposition, Reynolds's CEO and Chairman Bob Brockman testified that CDK's criticisms of Reynolds's data access policies were "harmful to [Reynolds] in the marketplace" and cost Reynolds hundreds of millions of dollars. (DRPSOF ¶ 25.)

19 CDK claims that it "started to change its marketing materials and dealer messaging to eliminate references to its more lenient data access policies as early as 2012"—for example, by removing its "Basic Access" program from marketing materials. (Defs.' Authenticom MSJ at 34.) But Plaintiffs dispute that those materials reflected a policy change; they note emails from CDK as late as 2015 advising dealers that they had the right to decide who accesses their data. (See PRDSOF ¶ 140.)

As Plaintiffs frame the record, "[t]he events of August 2013 illustrate the state of competition between the two DMS companies

prior to their conspiracy." (Authenticom Opp. at 14.) In August—a few months after Reynolds targeted Authenticom and SIS with a major lock-out initiative—Reynolds also began blocking access by DMI, CDK's subsidiary. In response, CDK set up a "war room" to apply pressure on Reynolds through Reynolds dealers. (DRPSOF ¶ 32.) Plaintiffs identify communications with dealers doing business with Reynolds in which CDK sales representatives criticized Reynolds's security measures. In one such communication, the CDK representative told a dealer that Reynolds "just want[s] to be able to charge you (and us, and our clients, even the OEMs) for something we'd given you at no cost"; in another, the CDK representative urged the dealer to switch to CDK. (Id.) In internal emails in August, Reynolds's top sales officials reported that their salespersons were being inundated with calls from dealers who were upset about Reynolds's lock-out efforts; these officials projected that September would be Reynolds's "poorest month in a couple years." (Id. ¶ 33.) As described earlier, Howard Gardner of CDK wrote that in the absence of a "negotiated outcome between" Reynolds and CDK, CDK would "go down in fire and flames rather than surrender the fight" and would face "Hard Landing" or "Rough Landing" scenarios in which Reynolds "continues to lock and relock user accounts," while Reynolds would "lose dealer contract renewals" as CDK "attacks [Reynolds] accounts." (Id. ¶ 34.) Thus, while Reynolds may have been technically able to block independent integrators, it did not make business sense to do so. After dealers, vendors, and third-party integrators pushed back, Reynolds whitelisted thousands of connections, exempting them from its blocking efforts (see id. ¶ 24), as it had done in the past for important clients and during major deal renewal periods. (See Section II.A.5, supra (detailing Reynolds's pre-conspiracy blocking efforts).)

Growing tension between CDK and Reynolds culminated in the September 27, 2013 meeting between Gardner and Schaefer. According to Gardner's summary of the meeting, Schaefer "opened up with the point that [Reynolds] would like to forge a common perspective with ADP on the topic of 'Data Security' messaging ... to synchronize our security perspectives and speak with similar voices to the industry." (PRDSOF ¶ 102.) Gardner believed that "this aspect of an agreement [wa]s very important" to Reynolds. (Id.)²⁰ In addition to "Market Messaging — Data Security," Gardner reported that CDK and Reynolds "conceptually" agreed on (1) reciprocal 3PA/RCI access; and (2) a "soft landing" for CDK's subsidiary integrators (DMI) via a whitelisting arrangement allowing DMI to continue offering its services on Reynolds's DMS. (Id.)

20 Indeed, Reynolds CEO Bob Brockman's June 2014 notes in preparation for a call with CDK CEO Steven Anenen make clear that Brockman himself was personally invested in seeing CDK moderate its tone. Brockman's notes reflect his position that CDK had "wrongly taken advantage of Reynolds in the marketplace over the issue of data security" and cost Reynolds "in the millions," which was "very much a personal" issue for Brockman. (See [1069-146].)

These records suggest that Reynolds wanted to reduce openness-based competition from CDK by encouraging CDK to embrace Reynolds's "security" perspective and close its system. CDK, in turn, seemed willing to change its (until-then-successful) way of doing business by closing its system, but only on the condition that DMI and its apps could continue serving Reynolds's dealerships. Plaintiffs characterize that arrangement as "a classic quid pro quo": "CDK agreed to abandon its competitive advantage in the DMS market by

refraining from marketing its open DMS" (and ultimately closing its DMS), and "[i]n return, Reynolds would give CDK's data integration businesses guaranteed access to the Reynolds DMS, exempting them from its blocking measures against other data integrators. And the parties would cooperate by giving each others' apps reciprocal access to their respective systems." (Authenticom Opp. at 17.)

Plaintiffs also cite evidence that, after the September meeting, CDK implemented an unwritten deal to stop competing on the issue of openness. Shortly after the meeting, Gardner instructed his team to stop criticizing Reynolds's closed system because of the companies' "new collaborative relationship," a point repeated internally over the coming months. (Id. at 18.) For instance, on October 8, a CDK sales executive declined an invitation to participate on a panel at an industry conference about Reynolds's "closed" policy and RCI's high prices, despite a recommendation from marketing staff that the panel could be used to "positively promote our dealer-friendly policies." (Id.) In declining, the executive said he wanted to "make sure any position we take publicly is consistent with the discussion at the end of last week." (Id.) In another example, Gardner declined to allow CDK support team members to share an article criticizing Reynolds's closed policy because it was "too inflammatory given the new collaborative spirit we're trying to develop with [Reynolds]" [975-56]. But Gardner said that the article "could come in handy" if CDK were unable to "reach an agreement" with Reynolds (id.), which according to Plaintiffs shows that CDK was "holding its competitive punches" because of the discussions with Reynolds. (Authenticom Opp. at 19.)

In addition, there is evidence that on several occasions, Gardner reassured Reynolds that, in promoting CDK's DMI data integration service, sales representatives would stop criticizing Reynolds's blocking measures. In October 2013, for example, Schaefer called and emailed Gardner to complain about a CDK sales representative who was criticizing Reynolds's blocking efforts. Schaefer said: "This is not the co-operation I thought we were in agreement"; Gardner then forwarded Schaefer's email internally and said: "we want to work on a more collaborative relationship." (See id. at 18-19.) In another, Gardner told Schaefer explicitly that the criticism was "unacceptable," and he would "take care of it." (Id.)

CDK spins that evidence, arguing that the communications only show it was wary of criticizing Reynolds while negotiating the DMI wind-down. (See Defs.' Authenticom MSJ at 36-37, 40-41.) Indeed, a jury may well be unmoved by these circumstances, but Plaintiffs' evidence is sufficient to defeat summary judgment. While CDK argues that much of Plaintiffs' evidence is unrelated to DMS competition over openness, there is sufficient evidence that is specific to openness. For example, Steve French, CDK's director of client and data services, sent a mass email to his DMI team in October 2013 remarking that "many OEMs" had started to "work[] directly with [Reynolds] to get data for their programs" [1089-167]. He noted "[i]t's DMI's position that we WANT and NEED to support these efforts and as such we are working very closely with [Reynolds] on doing so," and told salespeople that it was "very damaging to our cause if we say and write things about [Reynolds] that are not factual and that needs to STOP immediately. I cannot stress this enough!" (Id.) A reasonable jury could conclude that French was telling his team that CDK now supported Reynolds's practice of closing its DMS, or at least that it planned to stop criticizing that policy. Similarly, at the January 2014 NADA conference, CDK—which had previously touted open access to its system—shifted course, presenting a "Data Security Workshop" that warned of the risks of "giv[ing] people access to your DMS" and highlighted the benefits of CDK's data integration service 3PA, which offered "secured, monitored access." (Authenticom Opp. at 20.) CDK also sent letters to dealers and OEMs aimed to "combat" independent data integrators like Authenticom: it said Authenticom was not an approved vendor and suggested that Authenticom's practices could be unsafe. (See id. at 20-21; DRPSOF ¶ 40.) Around this same time, internal CDK emails suggested CDK management believed the "timing" was "good to move forward" with "secur[ing] our system" because "[Reynolds] is now succeeding in fully securing their system" and blowback would be minimal "especially under the cover of the new NADA security guidelines" (see Gardner Email, [1089-103]), that had been released in August 2013.

Viewed in the light most favorable to Plaintiffs, there is also evidence that CDK and Reynolds saw the 2015 Agreements as implementing a prior commitment to close their respective systems to data integrators. As Gardner described it, the Agreements would "secure [Reynolds's] DMS with the least possible disruption to their dealers" to "completely block Authenticom." (Authenticom Opp. at 24.) CDK viewed a "reciprocal agreement[]" with Reynolds as a way to "[s]ecure the CDK DMS through eliminating opportunities for key enablers of hostile integration." (Id. at 25.) In January 2015, shortly before reaching a final agreement, Gardner's notes stated clearly: "We have an agreement. We are locking down our DMS." (Id. (emphasis added).)

CDK devotes several pages of text to arguing that the 2015 Agreements do not explicitly require either CDK or Reynolds to block third-party access to their DMSs. (See Defs.' Authenticom MSJ at 3.) But as the court recognized at the motion to dismiss stage: "[s]uch a partial ceasefire and mutual forbearance between two rivals would make sense if—as the executives allegedly admitted—Defendants sought to 'support' one another's integration services to the exclusion of third-party integrators from the competition" and "it is reasonably inferable that such agreements were a part of the broader and admitted agreement to block third-party integrators." Authenticom MTD Op., 313 F. Supp. 3d at 951. The same holds at summary judgment; Plaintiffs have adduced evidence sufficient to require that a jury decide whether the 2015 Agreements were part of a broader conspiracy.

Plaintiffs also point to evidence that CDK took action to secure its DMS very soon after signing the 2015 Agreements. The timing supports a reasonable inference that CDK and Reynolds did more than just "agree[] to joint messaging"; indeed, a reasonable jury could find they "also conspired over their underlying data access policies." (Defs.' Authenticom MSJ at 37.) This is true even if "CDK began reevaluating its DMS access policy long before the start of the alleged conspiracy" (id. at 20)—and in any event, Plaintiffs dispute when that "reevaluation" took place. (See Section II.A.4, supra (reviewing pre-2013 CDK behavior).)

Five days after CDK signed the 2015 Agreements, Thorne instructed the research and development department to begin developing methods to "secure the platform ... GET THIS GOING IMMEDIATELY." (DRPSOF ¶ 60.) Securing its DMS became part of CDK's broader strategy known as "SecurityFirst," which it publicly announced in June 2015, but had revealed to some dealers as early as April.

(PRDSOF ¶¶ 157-58.) SecurityFirst, as CDK's expert Dr. Whinston characterized it, was CDK's signal that it "intended to secure its DMS" against independent integrators. (DRPSOF ¶ 60.) The timing of CDK's SecurityFirst initiative raises a question about whether CDK indeed "independently" secured its DMS. (See Defs.' Authenticom MSJ at 3-4; see also id. at 20-21.)

True, CDK did not close its DMS at the same time as Reynolds. But Reynolds was much closer to a fully "closed" system both in September 2013—when CDK began holding back criticism concerning Reynolds's "security" messaging—and in February 2015, when the formal agreements were signed and CDK was locked into the alleged agreement to "close" its DMS. And although CDK discounts other evidence of collusion as speculative, Plaintiffs have cited internal Reynolds emails and notes from 2014 reflecting that CDK had informed Reynolds that it planned to restrict independent access. (DRPSOF ¶ 49.) Plaintiffs also cite evidence that CDK gave Reynolds a preview of its security initiatives in 2016, before they were publicly revealed. (See Authenticom Opp. at 29.)

Perhaps most damning to CDK's position is Authenticom CEO Cottrell's testimony about the conversations he had with Defendants' executives in 2015 and 2016. Cottrell recalled a May 2015 phone conversation with Schaefer, shortly after Reynolds sent Authenticom a cease-and-desist letter. According to Cottrell, Schaefer told him that CDK and Reynolds "had an agreement to support each other's 3PA and RCI programs and therefore block competitors like Authenticom from pulling dealer data." (Id. at 30.) Defendants retort that "Cottrell did not believe" that "Schaefer had revealed a conspiracy" in that conversation. (Defs.' Authenticom MSJ at 45.) As Plaintiffs point out, though, Cottrell's contemporaneous belief about the agreement's legality says nothing about the fact of the agreement itself. (See Authenticom Opp. at 43-44.) CDK will have the opportunity to cross-examine Cottrell in front of the jury.

In a later conversation at the April 2016 NADA convention, Cottrell testified, CDK's McCray pulled Cottrell aside and told him that the "major DMS companies" were "working collaboratively to remove all hostile integrators from our DMS system." (Id. at 30.) McCray also alleged told Cottrell, "I'd really like to see you get something for [your company] before I f--king destroy it." (Id.) Defendants again insist this is not direct evidence of a conspiracy, because the phrase "major DMS companies" is "plainly ambiguous" and "one would have to infer (counterintuitively) that McCray used that roundabout phrase to refer to a bilateral conspiracy between CDK and Reynolds." (Defs.' Authenticom MSJ at 47-48.) But Cottrell testified that this is in fact what he understood McCray to mean (see Authenticom Opp. at 44), and that is entirely reasonable considering that CDK and Reynolds at that time controlled roughly seventy percent of the DMS market. (See DRPSOF ¶ 130.)

In ruling on the Defendants' motion to dismiss, the court found that the Cottrell conversations were "direct-evidence allegations of the agreement" to conspire because, "[t]aken as true, the fact that two of Defendants' executives admitted an agreement to block Authenticom and other integrators from accessing dealer data shows directly the existence of such an agreement. No further inference is required." Authenticom MTD Op., 313 F. Supp. 3d at 951. The court further determined that, even if "there was some ambiguity in the admissions" as Defendants argued, "they are at a minimum 'highly suggestive of the existence' [of an] agreement to block Authenticom

from accessing dealer data and thus [force Authenticom] out of the market." Id. (quoting High Fructose Corn Syrup, 295 F.3d at 663). The parties now dispute whether Cottrell's conversations are indeed "direct" evidence, but the court need not resolve the issue. Given the other circumstantial evidence, the economic evidence (discussed below), and Cottrell's conversations with Reynolds and CDK executives, a jury could reasonably conclude that CDK conspired with Reynolds. The jury may also conclude that, in CDK's words: "[w]hat Authenticom calls a conspiracy to stop competing on 'openness' simply reflects CDK's and Reynolds's separate decisions to secure their DMSs against hostile access at different times for their own interests." (Defs.' Authenticom MSJ at 19.) But that argument—like Plaintiffs'—must be presented to a jury.

b. CDK's remaining arguments

CDK has offered several additional arguments in support of its motion, but as explained below, none satisfy the court that summary judgment is appropriate.

Unilateral interests. First, CDK contends that Reynolds's and CDK's actions were fully consistent with each company's unilateral interests. As CDK notes, Plaintiffs' experts Drs. Israel and Lawton both testified that Authenticom's damages from unilateral action would be the same as its damages from collusive action. Accordingly, CDK asserts, "without a conspiracy, Reynolds and CDK could (and did) achieve their objectives of securing their respective systems" because they had the unilateral power to close their DMS and to do so profitably. (Defs.' Authenticom MSJ at 22.) But this argument rests on a constrained view of the record. Prior to the alleged conspiracy, CDK had no firm plan to close its DMS and an urgent need for Reynolds to stop blocking DMI's access. CDK also ignores the evidence provided by Cottrell. As Plaintiffs observe: a "reasonable jury could conclude that Defendants saw that they could both profit if they replaced competition on openness with collusion." (Authenticom Opp. at 54.)

Motive. CDK argues, further, that Plaintiffs' theory—under which "CDK immediately gave up a competitive advantage by ceasing to compete with Reynolds on 'openness' in exchange for speculative profits years in the future"—does not provide a sufficient incentive to conspire. (Defs.' Authenticom MSJ at 28.) To the contrary, CDK urges, it would not have agreed to incur near-term losses "'in exchange for the speculative possibility of more than making them up in the uncertain and perhaps remote future.'" (Id. (quoting High Fructose Corn Syrup, 295 F.3d at 661).) But CDK ignores other, more immediate benefits: the companies could avoid "fight[ing] it out in the DMS world for another 2 years or more," in Gardner's words. (Authenticom Opp. at 17.) CDK insists that Gardner's statement "has nothing to do with DMS competition" and that "the only reasonable interpretation of Gardner's statement is that DMI would continue 'fighting it out' with Reynolds to access the Reynolds DMS" if "the parties could not agree on an orderly wind down." (Defs.' Authenticom MSJ at 39-40.) But Gardner did not testify to that understanding; he said he did not remember what he meant. (DRPSOF ¶ 52.) Gardner's statement also reasonably supports a broader inference: "that the two rivals saw their agreement as superior to continued competition on openness in the DMS market that was becoming increasingly costly to both." (Authenticom Opp. at 52.) Both parties stood to gain from an agreement—Reynolds from CDK's commitment to stop trashing Reynolds's allegedly profit-motivated security as a charade and CDK from DMI's continued (and unique) access to

Reynolds's DMS. That is enough for a jury to find motive.

Enforcement mechanism. CDK next claims that "the absence of any evidence of enforcement [mechanism] undercuts the conspiracy claim." (Defs.' Authenticom MSJ at 31.) But the authority on which CDK relies recognizes that "a cartel may exist with only soft measures of control or ineffective enforcement" and that "evidence of an enforcement mechanism is not always required." Kleen Products, 910 F.3d at 937 (citation omitted). Plaintiffs have produced evidence of soft measures, like the parties' back-and-forth emails and calls when their employees strayed off message. Plaintiffs identify a more vigorous enforcement mechanism, as well: Reynolds's threat to resume blocking DMI and CDK's apps. (See Authenticom Opp. at 57-58.) CDK, in turn, held over Reynolds a threat to "fight it out in the DMS world" for several more years (id.)—a strategy that Reynolds's Brockman testified had already cost Reynolds hundreds of millions of dollars in lost revenues. (See DRPSOF ¶ 25.)

Switching data. According to CDK, data concerning dealer switching from one DMS to another is inconsistent with Plaintiffs' theory: "Reynolds won hundreds of dealers from CDK after 2013, and CDK won hundreds [of] dealers from Reynolds over the same period." (Defs.' Authenticom MSJ at 34.) But the relative level of competition is disputed. Plaintiffs respond that, although Reynolds and CDK "continued to win some business from each other" after 2013, "the share of Reynolds's competitive losses going to CDK declined despite the fact that Reynolds's blocking angered dealers—causing the worst month of DMS losses that Reynolds had ever experienced in the month before Defendants' agreement." (Authenticom Opp. at 59.)

Authenticom connections. CDK observes that Authenticom's connections to the Reynolds DMS are higher today than in 2013. But there is a material dispute over the proper inference to draw from that data. According to Authenticom's expert (Lawton), the growth rate of those connections declined dramatically after September 2013 and that decline worsened after the 2015 Agreements were signed. In addition, Reynolds dealers must manually "push" data to Authenticom—a method that is arguably inferior to automated integration. (See id. at 59-60.) Connections to CDK raise another factual dispute: according to Plaintiffs, the evidence shows that the growth of Authenticom's CDK connections slowed in the fall of 2013 after CDK and Reynolds began colluding, fell immediately after the 2015 Agreements, and decreased even more sharply in 2016 after CDK began blocking. (Id. at 60.)

* * *

For the reasons discussed above, CDK is not entitled to summary judgment on the "agreement" element of Plaintiffs' Sherman Act § 1 conspiracy claims. Even if "no single item of evidence points unequivocally" to conspiracy—and here, there is some arguably "direct' evidence—the record, viewed in the light most favorable to Plaintiffs, is sufficient to survive summary judgment. High Fructose Corn Syrup, 295 F.3d at 655.

2. Element 2: Unreasonable restraint

The second element of Plaintiffs' § 1 claim—an unreasonable restraint on trade—hinges on whether the alleged violation is per se unreasonable or is instead subject to the "rule of reason." See In re Sulfuric Acid Antitrust Litig., 743 F. Supp. 2d 827, 864 (N.D. Ill. 2010). A per se violation is presumptively unreasonable. Authenticom MTD Op., 313 F. Supp. 3d at 950 (quotations, citations, and brackets omitted). Rule-of-reason analysis, on the

other hand, requires that a plaintiff plead "anticompetitive effects" and "that the injury complained of be of a type that the antitrust laws were designed to guard against, and further that the antitrust violation be the direct cause of plaintiff's injury." Havoco of Am., Ltd. v. Shell Oil Co., 626 F.2d 549, 556 (7th Cir. 1980) (internal citation omitted). A three-step, burden-shifting framework determines whether a restraint violates the rule of reason. The plaintiff must first show that a challenged restraint has a "substantial anticompetitive effect that harms consumers in the relevant market." Ohio v. Am. Express Co., 585 U.S. 529, 138 S. Ct. 2274, 2284, 201 L.Ed.2d 678 (2018). Then, the defendant must "show a procompetitive rationale for the restraint." Id. If the defendant does so, "the burden shifts back to the plaintiff to demonstrate that the procompetitive efficiencies could be reasonably achieved through less anticompetitive means." Id.

CDK argues that rule-of-reason analysis applies here because: 1) the alleged conspiracy is unlike any of the traditional per se categories; 2) there are a number of procompetitive benefits to the 2015 Agreements; 3) the alleged conspiracy involves vertical as well as horizontal elements; 4) the conspiracy implicates federal laws like the CFAA and DMCA; and 5) the court must take into account negative externalities of openness, like DMS data breaches or performance disruption. Though the issues are thorny, the court agrees that rule-of-reason analysis applies.

First, as CDK points out, the alleged conspiracy bears little resemblance to agreements typically given per se treatment, like "'price-fixing schemes, as well as certain types of group boycotts, market allocations, and tying arrangements.'" (Defs.' Authenticom MSJ at 52-53 (quotation omitted).) Of course, schemes need not "fit precisely the characterization of a prototypical per se practice" to merit per se treatment. United States v. Andreas, 216 F.3d 645, 667 (7th Cir. 2000). One such practice is a group boycott, but "[e]xactly what types of activity fall within" that "forbidden category is ... far from certain." Nw. Wholesale Stationers, Inc. v. Pac. Stationery and Printing Co., 472 U.S. 284, 294, 105 S.Ct. 2613, 86 L.Ed.2d 202 (1985).

At the motion-to-dismiss stage, the court concluded that Plaintiffs had plausibly alleged a horizontal group boycott in Defendants' alleged agreement to block Authenticom and other integrators from accessing dealer data. As the court explained, "the Complaint in essence claims that Defendants conspired to block integrators from accessing necessary data from the dealers—meaning, 'relationships' integrators 'need'—by (among other things) withholding authorization and disabling third-party credentials, so that Defendants, ultimately, could charge more for their integration services." See Authenticom MTD Op., 313 F. Supp. 3d at 952.

In their reply brief (Defs.' Authenticom Reply at 26), Defendants recognize that under Northwest Wholesale Stationers, a cooperative venture's "concerted refusal to deal with [a competing firm] on substantially equal terms ... might justify per se invalidation if it placed a competing firm at a severe competitive disadvantage." See 472 U.S. at 295 n.6, 105 S.Ct. 2613 (quotation omitted). But they maintain that the record cannot support a finding that Plaintiffs (and particularly Authenticom) were put at a severe competitive disadvantage by the alleged conspiracy because "Authenticom's witnesses concede that manual push is a 'painless' process" and because Authenticom's business with both CDK and Reynolds dealers grew significantly during the supposed conspiracy. (Defs.' Authenticom Reply at 26.) Those facts, however, are disputed: testimony from employees

of CDK, Reynolds, Authenticom and other integrators, and of dealers suggests that manual push is not adequate and not desirable. (See PRDSOF ¶ 40.) As discussed above, the data concerning connections is also disputed. See Section II. B.1.b, supra. A reasonable jury could find that the alleged conspiracy put Authenticom and Plaintiffs at "a severe competitive disadvantage." Nw. Wholesale Stationers, 472 U.S. at 295 n.6, 105 S.Ct. 2613.

That said, the parties appear to agree that the 2015 Agreements—vertical agreements giving CDK access to data on the Reynolds DMS and vice-versa, in contrast to the overall horizontal conspiracy concerning DMS "openness"—must be analyzed under the rule of reason, like "nearly every other vertical restraint." Am. Express, 138 S. Ct. at 2284. "The purpose of the rule is to distinguish between restraints with anticompetitive effect that are harmful to the consumer and restraints stimulating competition that are in the consumer's best interest." Marion Diagnostic Ctr., LLC v. Becton Dickinson & Co., 29 F.4th 337, 348 (7th Cir. 2022) (quotation omitted). In Plaintiffs' view, rule-of-reason analysis will matter only if a jury concludes that their claims are limited to a challenge to the 2015 Agreements—in other words, only if the jury concludes that CDK and Reynolds never conspired to close their DMSs. (See Authenticom Opp. at 63-64.) But because the jury may well reach that conclusion, Plaintiffs will put on a full rule-of-reason case either way, undercutting the "litigation efficiency" that partly justifies the per se rule. In re Outpatient Med. Ctr. Emp. Antitrust Litig., 630 F.Supp.3d 968, 986 (N.D. Ill. 2022).

Moreover, the Supreme Court has warned against "indiscriminate" expansion of the circumstances that support per se analysis, especially "in the context of business relationships where the economic impact of certain practices is not immediately obvious." F.T.C. v. Ind. Fed'n of Dentists, 476 U.S. 447, 458-59, 106 S.Ct. 2009, 90 L.Ed.2d 445 (1986). In group boycott cases that merited per se review, the challenged practices were not plausibly "intended to enhance overall efficiency and make markets more competitive." Nw. Wholesale Stationers, 472 U.S. at 294, 105 S.Ct. 2613. The Seventh Circuit likewise requires plaintiffs to meet a high bar to trigger per se analysis. Plaintiffs must show: "(1) the boycotting firm has cut off access to a supply, facility or market necessary for the boycotted firm ... to compete; (2) the boycotting firm possesses a 'dominant' position in the market (where 'dominant' is an undefined term, but plainly chosen to stand for something different from antitrust's term of art 'monopoly'); and (3) the boycott ... cannot be justified by plausible arguments that it was designed to enhance overall efficiency." Toys "R" Us, 221 F.3d at 936.

Those elements are difficult for Plaintiffs to meet here, where they have alleged a complex conspiracy (with substantial vertical elements) between DMS providers, aimed at enabling them to raise prices in the separate but related data integration market. (See Reynolds's MVSC MSJ [955] at 9 n.10 (making this argument against MVSC).) Further, CDK plausibly argues that at least the 2015 Agreements—which were "integral" to "codifying" the conspiracy (Defs.' Authenticom MSJ at 53)—are procompetitive: The orderly wind-down allowed DMI to continue serving dealers and vendors without interruption for set periods. The RCI and 3PA Agreements also guaranteed CDK and Reynolds secure access to the other's DMS, broadening distribution of each firm's apps. Indeed, Plaintiffs' expert, Dr. Israel, recognized that the 2015 Agreements contained at least some provisions that "could be procompetitive." (Defs.' Authenticom Reply at 27.) For these reasons, the court heeds the Seventh Circuit's advice that "[i]t is a bad idea to subject a novel way of doing business (or an old way in a new and previously unexamined context ...) to per se treatment under antitrust law." In re Sulfuric Acid Antitrust Litig., 703 F.3d 1004, 1011 (7th Cir. 2012) (citation omitted). "The per se rule is designed for cases in which experience has convinced the judiciary that a particular type of business practice has no (or trivial) redeeming benefits ever," id. at 1011-12—a high bar that has not been cleared in this case. The court will instruct the jury on rule-of-reason analysis at trial.

3. Element 3: Injury

"To state an antitrust injury, a plaintiff must allege an anticompetitive injury that flows from defendant's actions and that the antitrust laws were intended to prevent." Chi. Studio Rental, Inc. v. Ill. Dep't of Com, 940 F.3d 971, 978 (7th Cir. 2019) (citation omitted). This requirement "ensures that a plaintiff can recover only if the loss stems from a competition-reducing aspect or effect of the defendant's behavior," id., by "filter[ing] out complaints by competitors and others who may be hurt by productive efficiencies, higher output, and lower prices, all of which the antitrust laws are designed to encourage." U.S. Gypsum Co. v. Ind. Gas. Co., 350 F.3d 623, 627 (7th Cir. 2003).

CDK argues that Plaintiffs cannot establish injury because so-called hostile access to its DMS violated the CFAA, the Copyright Act, and the DMCA. In ruling on the motions to dismiss, the court explained that "[e]ven assuming, without deciding, that the CFAA" and other cybercrime laws "prohibit[] Authenticom from accessing DMSs without Defendants' authorization, Defendants are not free to withhold such approval pursuant to illegal arrangements." Authenticom MTD Op., 313 F. Supp. 3d at 948 (citing cases); see also Ind. Fed'n of Dentists, 476 U.S. at 465, 106 S.Ct. 2009 ("That a particular practice may be unlawful is not, in itself, a sufficient justification for collusion among competitors to prevent it."). At the time, the court found that the alleged conspiracy plausibly "blocked or foreclosed Authenticom from the data-integration market"—in other words, "that Authenticom lacks authorization or access to Defendants' DMSs (which is necessary to compete in the data-integration market, given Defendants' upstream duopoly) because of Defendants' anticompetitive conduct." Authenticom MTD Op., 313 F. Supp. 3d at 948.

Plaintiffs' antitrust injury turns on whether the "desired trade" of independent integrators like Authenticom—i.e., automatic retrieval of data from CDK's DMS—was "illegal independent of the defendants' complained-of conduct." Authenticom MTD Op., 313 F. Supp. 3d at 947. In other words, the court first considers the legitimacy of Authenticom's conduct without regard to any conspiracy. CDK takes the position that "independent of 'authorization,' Authenticom's access to both the CDK and Reynolds DMS infringed Defendants' copyrights by creating unauthorized copies of copyrighted material." (Defs.' Authenticom Reply at 30.) But the Copyright Act grants the copyright holder the right to decide who may copy its software and related works. See 17 U.S.C. § 100 et seq. In other words, CDK can provide, or revoke, authorization to create such copies by accessing its DMS. The same is true of the CFAA, which prohibits "intentionally access[ing] a computer without authorization or exceed[ing] authorized access." 18 U.S.C. § 1030(a)(2); id. § (a)(5)(C). In other words, CDK's position assumes that CDK had not granted Authenticom those kinds of permissions.

And because there is a material factual dispute about whether Authenticom was authorized to access CDK's DMS at

the time the alleged conspiracy began, CDK is not entitled to summary judgment on this basis. True, CDK's standard Master Services Agreement has long prohibited dealers from allowing access to their DMS "by any third parties except as otherwise permitted by this agreement." (PRDSOF ¶ 67.) But there is extensive evidence that prior to the beginning of the alleged conspiracy, CDK did not rely on that provision to justify excluding independent integrators from access to its DMS. To the contrary, CDK marketed "Basic Access"—which permitted dealers to "provide[] access to third-part[ies] through dealer supplied User IDs and strong passwords as well as dealer provided modem access" while "maintain[ing] all responsibility for system and data access." (Id. ¶ 140.) CDK claims it dropped Basic Access by October 2012, prior to the start of the conspiracy. This, too, is disputed. CDK cites a 2012 marketing publication that no longer mentioned it, but Plaintiffs say that the publication does not definitively list all third-party access methods or show that CDK had affirmatively revoked independent access as of 2012. Indeed, Plaintiffs identify emails from CDK as late as 2015 advising dealers that they were free to decide who may access their systems. See Section II.A.4, supra (discussing this evidence and the parties' positions).

All of CDK's remaining evidence on the issue of unauthorized access post-dates the start of the alleged conspiracy. First, CDK advised dealers in August 2014 that Authenticom's access was unauthorized. Second, in April 2016, a CDK employee told Authenticom's CEO that CDK's dealer contracts do "not allow them to give you a user name and password to continue to export data," and that CDK would "sue [dealers] if needed to keep you out of our systems." (Defs.' Authenticom MSJ at 62.) Third, in March 2016, CDK began deploying a prompt on Authenticom accounts stating that only "authorized dealer employees" had permission to access CDK's system. (Id.) In light of the timing, these circumstances do not by themselves defeat Plaintiffs' claim that the alleged conspiracy caused injury.

4. Damages

To recover under the antitrust laws, a plaintiff "must prove that [its] damages were caused by the unlawful acts of the defendant." MCI Commc'ns Corp. v. Am. Tel. & Tel. Co., 708 F.2d 1081, 1161 (7th Cir. 1983). Once causation is established, "[c]alculating damages presents a separate challenge." Sulfuric Acid, 743 F. Supp. 2d at 866. "'Damages calculations in antitrust cases seek to compare plaintiffs' actual experience in the real world with what the plaintiffs' experience would have been, 'but for' the antitrust violation.'" In re Rail Freight Fuel Surcharge Antitrust Litig., 292 F. Supp. 3d 14, 56 (D.D.C. 2017) (quotation omitted). "[T]he amount of damages may be determined by a just and reasonable estimate" but cannot be "the product of speculation or guess work." MCI, 708 F.2d at 1161. In complex cases involving "a series of unlawful acts intertwined with one another," courts do not require "strict proof of what damages have been caused by which acts," a standard that "serves to prevent a defendant from profiting from his own wrongdoing." Id. (citing Bigelow v. RKO Radio Pictures, 327 U.S. 251, 264-65, 66 S.Ct. 574, 90 L.Ed. 652 (1946)). "It is essential, however, that damages reflect only the losses directly attributable to unlawful competition." Id. "If a plaintiff has suffered financial loss from the lawful activities of a competitor, then no damages may be recovered under the antitrust laws." Id.

a. Dealer Plaintiffs

In its Daubert ruling, the court excluded the opinion of Dr. Williams (the Dealers' expert) on federal antitrust damages because

his analysis included integration overcharges that were passed on to the Dealers by vendors. The court explained that under Illinois Brick Co. v. Illinois, 431 U.S. 720, 97 S.Ct. 2061, 52 L.Ed.2d 707 (1977), and Apple, Inc. v. Pepper, — U.S. —, 139 S. Ct. 1514, 203 L.Ed.2d 802 (2019), the Dealers could seek to recover damages as direct purchasers of DMS but could not—at least under federal law—claim damages stemming from an "overcharge" passed "down a chain of distribution." Apple, 139 S. Ct. at 1525. The court expressed no opinion on whether Williams could offer the same opinion in support of a claim for damages under the antitrust laws of one or more states, as the parties had not adequately briefed that issue. See Daubert Op., 581 F. Supp. 3d at 1065.

Adhering to that earlier ruling, the court grants summary judgment against the Dealers on their § 1 horizontal conspiracy claim for damages (Count I), though their claim for injunctive relief (Count II) survives. The Dealers have no admissible evidence that they suffered direct damages as purchasers of DMS. Their only evidence of damages is Dr. Williams's report, which is based solely on indirect damages—i.e., the alleged integration overcharges that were passed on from vendors to the Dealers. (See CDK Dealers MSJ at 15.)

b. AutoLoop

i. Dr. Israel

CDK contends that AutoLoop's damages are inflated in at least three ways, as discussed more fully below. CDK's principal challenge to the damages claim, however, is an attack on AutoLoop's expert, Dr. Israel.

Dr. Israel observed that Defendants competed on DMS "openness" for many years. (See Pls.' Daubert Br. [1001] at 8.) Before the alleged collusion, Reynolds more vigorously blocked access by independent integrators, whereas CDK marketed that its dealers were free to use independent integrators on its platform. Then, in September 2013—shortly after Reynolds began taking more aggressive blocking action—CDK abruptly stopped competing with Reynolds on "openness." (Id. at 10 (citing Israel Rep. ¶¶ 153-54).)

To measure the impact of that alleged collusion on data integration prices, Dr. Israel examined data integration prices from January 1, 2013 to early 2019. He quantified damages using a "difference-in-differences" model, comparing price increases for RCI and 3PA with price increases for a benchmark of other data integrators during the same period. (See AutoLoop MSJ Opp. [1075] at 1.) In other words, Israel's model calculates damages from unlawful conduct by measuring how much 3PA and RCI prices increased as compared to the benchmark prices of SIS and Authenticom, which are not alleged to have conspired and which, according to Dr. Israel, would have been subject to the same market forces. (See id. at 6-7.) Israel conducted robustness checks using pre-conspiracy 3PA and RCI prices as an alternative benchmark and reached a similar calculation under that approach. (Id.)

CDK's summary judgment arguments largely mirror its earlier Daubert challenge to Dr. Israel's damages model, which the court rejected. See Daubert Op., 581 F. Supp. 3d at 1050-57. The summary judgment inquiry is slightly different, however. Daubert asks whether Israel was qualified, whether his methodology was scientifically reliable, and whether his testimony would assist the trier of fact to understand the evidence or determine a fact in issue. See Myers v. Ill. Cent. R.R. Co., 629 F.3d 639, 644 (7th Cir. 2010). Summary judgment requires the court to evaluate the whole record to determine whether "the evidence is such that a reasonable jury could return

a verdict for the nonmoving party." Liberty Lobby, 477 U.S. at 248, 106 S.Ct. 2505.

CDK argues that it is entitled to summary judgment because Dr. Israel's analysis (that is, AutoLoop's damages theory) did not measure the harm caused solely by Defendants' allegedly unlawful conduct. Instead, CDK claims, Israel improperly attributes nearly all of 3PA's price increases starting in September 2013 to the alleged antitrust violation. (See CDK AutoLoop MSJ at 6.) First, CDK faults Israel for failing to consider "the many reasons why firms might raise prices unilaterally due to market factors," including "growing demand for certified integration; the increased costs of building-out certified integration programs or responding to growing cyber-security threats; and a growing need for improved or sophisticated integration services." (Id. At 7.) According to Defendants, these "factors would not affect demand for Authenticom or SIS" because they are not DMS providers and do not offer certified integration. (Id.)

That argument could have traction, but CDK fails to meaningfully develop it. It does not, for example, point to any evidence that its status as a certified integrator justified price increases or that price increases coincided with increases in product quality. CDK's expert, Dr. Murphy, echoes that "there are other important factors affecting 3PA prices" starting in September 2013, but offers no specific factual support for that opinion. (See Murphy Rep. [975-76] at 44-45.) There are surely many reasons why CDK could have increased prices, but CDK's speculations on this score do not so thoroughly undermine Dr. Israel's report that no jury could credit it. To the contrary, a jury could reasonably conclude that any supposed improvements to 3PA were not valued by app vendors or did not exist and thus had no effect on 3PA's prices.

Second, CDK argues that Dr. Israel's damages model does not account for CDK's significant unilateral power. But Israel explained how his "differences-in-differences" model accounts for unilateral market power by comparing 3PA's prices against various benchmarks before and after the conspiracy. According to Israel, those benchmarks isolate only those price increases that cannot be explained by unilateral market power. And since there is no evidence that CDK and Reynolds achieved greater market power during the damages period—or at least power unrelated to the unlawful agreement—there is no basis to credit unilateral power for price increases during the alleged conspiracy period.

Unsatisfied with that explanation, CDK urges that "Dr. Israel's conclusion that there would be at most negligible price increases in the but-for world" is "undermined" by his finding that "CDK and Reynolds each had the power to raise integration prices unilaterally." (CDK AutoLoop MSJ at 7.) According to CDK, Israel opined that CDK could exercise this power "profitably" and "his damages calculation for the unilateral claims is identical to his calculation for the conspiracy claims." (Id. at 8.) In other words, as CDK reads his report, Israel concluded that "prices increase the same with or without the alleged conspiracy." (Id.)

Several weaknesses in that logical chain preclude summary judgment. As an initial matter, CDK overstates what Dr. Israel says about its ability to raise integration prices unilaterally. Although Israel wrote that CDK had "sufficient market power in the DMS market to be able to profitably implement a policy of blocking independent third-party [integration] providers," he also said the policy was "substantially more profitable when implemented through concerted action with its principal

DMS market rival." (Israel Rep. ¶ 211; see also id. ¶ 110 (explaining how his report will "describe how competition between CDK and Reynolds (on openness in particular) hindered each company's unilateral ability to block independent [integration] providers' access to dealer data and thus hindered their ability to leverage fully their joint market power").)

The court also disagrees that Dr. Israel effectively admitted that "prices increase the same with or without the alleged conspiracy" when he calculated the same damages for AutoLoop's unilateral claims as for the conspiracy claim. (See CDK AutoLoop MSJ at 8.) Dr. Israel explained that it was "clear" to him "as a matter of economics" that CDK and Reynolds "communicated, reached explicit agreements, and took a series of actions together to leverage their joint market power," but he clarified that he was not opining about whether their actions "satisfy the legal standards" for a § 1 conspiracy. (Israel Rep. ¶ 205.) With respect to the "unilateral claims"—based on exclusive dealing or monopolization, rather than collusion—Israel concluded "that very little of the economic analysis changes" because "[a]ll of the actions that CDK and Reynolds took ... to monopolize [data integration services] and harm App competition continue to be true." (Id. ¶ 206.) As AutoLoop puts it, Dr. Israel's "point about unilateral conduct was that even if CDK's and Reynolds's conduct after September 2013 was found to be unlawful unilateral action rather than unlawful collusive action" damages remain the same because "the effects of the unlawful conduct" do not depend on the legal basis for liability. (AutoLoop MSJ Opp. at 11.) Put simply, Israel's position is that CDK did what it did, and AutoLoop suffered as a result. Of course, a jury might find that CDK's actions were both unilateral and lawful, but the court cannot make that finding at summary judgment.

Contrary to CDK's characterization, Dr. Israel also did not opine that "there would be large price increases absent a conspiracy." (CDK AutoLoop MSJ at 8.) At Israel's deposition, defense counsel questioned him extensively about CDK's "unilateral interest" in the "but-for" world. (Israel Dep. [979-122].) Israel cautioned that this is "a term that can mean many things in many contexts" and "is often greatly misused and misunderstood in ... economics." (Id. at 59:1-10.) Dr. Israel explained that although in a "pro and con" list there might be some "pros" to CDK's closing its system—mainly, that CDK's certified integration business could charge higher prices because it would face less competition—he saw no evidence that on balance CDK would have closed its system absent a conspiracy. (Id. at 63:11-18, 67:15-69:1, 255:6-259:19.) He also reasoned from CDK's history: over the years, Israel observed, "as Reynolds became more closed, CDK stayed open and competed as a DMS provider on that basis"; and CDK "never closed its system until it cooperated with Reynolds." (Id. at 72:3-73:7.) Doing so—that is, closing its system—would have rendered CDK "a less effective DMS competitor" absent a conspiracy. (Id. at 75:1-21.) Israel acknowledged that CDK could have made a unilateral decision to close its system, but explained that "just because you would have some ability on your own to raise the price of your 3PA doesn't tell you that that would outweigh those other effects," or "what the mark of equilibrium would be if you tried to do that on your own." (Id. at 75:22-76:5; see also Israel Rep. ¶ 12 n.6 (explaining that although "CDK had the power to raise [integration] prices and revenues, it determined that it was better off marketing its more open system to take DMS business from Reynolds until reaching agreement with Reynolds to cooperate, eventually leading to the parties' jointly monopolizing their [integration] services"). Put differently, Israel

opined that CDK would have lost more money from lower DMS market share market than it made by raising data integration prices.

CDK will be free to cross-examine Dr. Israel on these issues at trial, but its challenges to his damages opinion are insufficient to merit summary judgment.

ii. Damages attributable to Cox

AutoLoop seeks damages on behalf of a putative class, which includes eight subsidiaries of Cox Automotive, Inc.: Autotrader.com, Inc.; Dealer Dot Com, Inc.; DealerTrack, Inc.; HomeNet, Inc.; Kelley Bluebook, Co., Inc.; vAuto, Inc.; VinSolutions, Inc.; and Xtime, Inc. (collectively the "Cox Vendors"). In 2019, Cox and CDK settled their claims against one in another in this case [723], which CDK argues precludes them from recovering damages in this suit. AutoLoop, as putative class representative, represents that it will not seek damages based on harm suffered by Cox. (See AutoLoop MSJ Opp. at 15.) The court therefore grants summary judgment in favor of CDK on this issue and will direct Dr. Israel to omit damages attributable to Cox and its subsidiaries for the jury.

iii. Reynolds's Prices

As it did in its Daubert motion, CDK argues that Dr. Israel's overcharge analysis improperly uses average revenue per unit ("ARPU") to measure the impact of Defendants' alleged collusion on RCI's prices. According to CDK, that methodology improperly fails to account for possible "composition effect"—a shift by app vendors toward more expensive integration packages rather than an increase in the prices for the existing packages. Accordingly, CDK argues that it cannot be liable for any damages AutoLoop seeks to recover for RCI overcharges. (CDK AutoLoop MSJ at 13-14.) Dr. Israel, for his part, explains that ARPU remains the correct measure for an overcharge analysis because it measures "actual marketplace prices," i.e., the total amount of data integration fees paid by an app vendor. (See AutoLoop MSJ Opp. at 11-12.)

At the Daubert stage, the court reasoned that the parties' experts advanced competing opinions about how best to calculate damages attributable to data integration price increases, and it was not the court's role to evaluate the quality of their data, inputs, and conclusions. Daubert Op., 581 F. Supp. 3d at 1054. The court also found that unlike the defendants in the cases cited by CDK, nothing here suggests "that a hypothetical composition effects error has skewed Israel's calculations in any material way." Id.

CDK's "composition effect" argument fares no better at summary judgment. Dr. Israel calculated RCI prices per app, meaning that composition effects would come into play only if an app changed the mix of integration packages that it used over time. Israel believed it was unlikely that the integration needs of an app would change over time; CDK presents no evidence that his hypothesis was wrong. To test his hypothesis, Israel performed a "robustness check" using a "fixed weight price index" (AutoLoop MSJ Opp. at 12-13)—an average price "taken over a fixed mix of products, thus eliminating changes in product mix over time." (Israel Reply Rep. [957-2] ¶ 194.) He compared the fixed weight price index and ARPU and found that the two measures of RCI prices were "nearly identical," which, he concluded, suggests that composition effects are not driving the price increases he observed. (Id.) CDK has not satisfied the court that its concern about composition effects precludes calculation of damages to a reasonable degree of certainty.

iv. Forced-switching damages

Dr. Israel calculated damages incurred by vendors who demonstrated a

preference for lower-priced integrators SIS and Authenticom but were forced to switch to certified integration through RCI or 3PA. This group excluded vendors who did not use independent integrators or who already used certified integration—i.e., accessed Reynolds's DMS via RCI or CDK's DMS via 3PA—for most of their connections. (See CDK AutoLoop MSJ at 14; AutoLoop MSJ Opp. at 13.) Israel began his analysis by assuming that the remaining vendors valued certified integration and independent access equally, and thus were indifferent as between the two options. CDK challenges this assumption, and contends it was error for Israel to disregard the additional value of certified integration, which is "essential ... to provide the highest level of product quality." (CDK AutoLoop MSJ at 16.) But Israel provided a reasonable explanation and factual basis for his assumption: evidence of app vendors who believed independent integrators offered a superior product or who saw no difference in the quality of services provided by 3PA after significant price increases. (Israel Reply Rep. ¶¶ 150-54.) AutoLoop also argues that, because Israel calculated "forced switching" damages only for app providers who were already mostly using SIS and Authenticom, those vendors were more likely to either be indifferent to quality or to prefer independent integration, meaning Israel's analysis was "well-founded and even conservative." (AutoLoop MSJ Opp. at 14.) CDK may be correct that certified integration is more expensive "because it provides vendors more functionality" (CDK AutoLoop Reply [1122] at 10), but this is something that CDK can explore by way of cross-examination.

C. Other Sherman Act Claims

In addition to horizontal conspiracy claims, Plaintiffs bring Sherman Act claims for (1) exclusive dealing under § 1; and (2) monopolization of CDK's and Reynolds's separate data integration services "aftermarkets" under § 2. Both claims are rooted in the theory that CDK and Reynolds allegedly required vendors seeking to participate in the 3PA or RCI programs to purchase "data integration services" and receive data exclusively from CDK or Reynolds.

1. Section 1 exclusive dealing

AutoLoop and the Dealers allege that CDK's vendor contracts include unlawfully anticompetitive exclusive dealing provisions. (AutoLoop Compl. ¶¶ 182-89; Dealers Compl. ¶¶ 190-199.)²¹ In January 2016, AutoLoop entered into a Managed Interface Agreement with CDK, in which AutoLoop agreed to access data stored on CDK's DMS only through 3PA. Specifically, Section 1(f) of the Agreement prohibits AutoLoop from receiving "any" data sources from the CDK DMS through any source outside of the 3PA program, including "any" third party.²² (Defs.' Joint Statement

21 AutoLoop's claim is for damages and injunctive relief. The Dealers' sole remaining claim is for injunctive relief. Dealers MTD Op., 362 F. Supp. 3d at 530-31 (dismissing damages claim under Illinois Brick).

22 Section 1(f) states: "Vendor agrees to use its best efforts so that ... all Vendor Clients are using the Applications exclusively through the Managed Interface System in the manner described on the relevant [Statements of Work] by no later than the one (1) year anniversary of the date of this Agreement. After such one (l) year anniversary of the date of this Agreement, ... Vendor agrees that it will not (a) otherwise access, retrieve, license, or otherwise transfer any data from or to an [sic] CDK System (including, without limitation, pursuant to any 'hostile interface') for itself or any other entity or (b) contract with, or otherwise engage, any third party (including any Vendor Client) to access, retrieve, license or otherwise transfer any data from or to an [sic] CDK System[.]" (Defs.' JSOAF ¶ 114.)

of Additional Material Facts ("JSOAF") [1062] ¶ 114.)

"An exclusive dealing contract obliges a firm to obtain its inputs from a single source." Paddock Publ'ns, Inc. v. Chi. Tribune Co., 103 F.3d 42, 46 (7th Cir. 1996). In this case, Section 1(f) of the Agreement requires AutoLoop to obtain data-integration services through 3PA alone. In some circumstances, exclusive dealing agreements have procompetitive benefits. For example, they can increase efficiency and prevent free-riding. But such provisions are objectionable if they unreasonably "deny outlets to a competitor during the term of the agreement." Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380, 393 (7th Cir. 1984). Accordingly, exclusive dealing "must be analyzed" under the rule of reason. Dos Santos v. Columbus-Cuneo-Cabrini Med. Ctr., 684 F.2d 1346, 1352 (7th Cir. 1982). At the first stage of the rule-of-reason analysis, a plaintiff must show that: (1) the exclusive dealing "foreclose[s] competition in a substantial share of the line of commerce at issue," Republic Tobacco Co. v. N. Atl. Trading Co., 381 F.3d 717, 737-38 (7th Cir. 2004), resulting in "at least one significant competitor of the defendant" who is excluded "from doing business in a relevant market"; and (2) "the probable (not certain) effect of the exclusion" must be to "raise prices above (and therefore reduce output below) the competitive level, or otherwise injure competition." Roland Mach., 749 F.2d at 394

At the motion to dismiss stage, the court held that Plaintiffs' "allegations raise a reasonable inference that, if enforced, Defendants' exclusive-dealing provisions with vendors could foreclose a substantial portion of the data-integration market and reduce output." Authenticom MTD Op., 313 F. Supp. 3d at 958. The court noted that Defendants' arguments were "ill-timed" for a motion to dismiss. That changes at this stage; to defeat summary judgment, Plaintiffs must now bring forth evidence from which a reasonable jury could find that the exclusive dealing provisions foreclosed a substantial portion of the data integration market.

As part of this inquiry, the party challenging an exclusive-dealing agreement court must find that a competitor was excluded from "a relevant market" in terms of product and geography. Roland Mach., 749 F.2d at 394. CDK argues that Plaintiffs have not properly defined the market, but the court rejected that argument in its Daubert opinion. See Daubert Op., 581 F. Supp. 3d at 1056 (definition of relevant market is a battle of the experts best left to the factfinder). For purposes of summary judgment, the court accepts Plaintiffs' proposed definition—a distinct nationwide data integration aftermarket for "vendors serving CDK dealers." Id. The remaining question is whether Plaintiffs have adduced evidence that the exclusive dealing provisions themselves foreclosed Authenticom from that market. As explained below, the court concludes they have not.

AutoLoop's evidence on this issue comes from its expert, Dr. Israel. But Israel's opinions about the effects of CDK's exclusive dealing provisions appear in a single paragraph, where he describes the provisions and notes that, even if a vendor "were to find some way to use an independent [integrator] on a CDK or Reynolds DMS" for one of their apps, the vendor "would be in violation of its exclusive contract." (Israel Rep. ¶ 159.) In his reply report, Israel again mentions exclusive dealing only once, in a footnote where he defines "blocking" as a "suite of behavior" including technological tactics and "efforts to enforce exclusive dealing." (Israel Reply Rep. ¶ 5 n.3.) Fatally for AutoLoop, Israel makes no effort to isolate the anticompetitive

effects of the exclusive dealing provisions. That means AutoLoop has no evidence that the exclusive dealing provisions—rather than or in addition to CDK's technological blocking efforts—caused Authenticom's foreclosure from the integration market. Israel does not suggest that CDK could only have secured—or even more effectively secured—its DMS by way of the exclusive dealing provisions in its contracts with vendors. Put another way: start from the assumption that CDK engaged in technological blocking, which it did. Regardless of whether that blocking was lawful or unlawful, and whether it was unilateral or collusive, AutoLoop has no evidence that the exclusive dealing provisions themselves had any marginal effect on top of CDK's other blocking efforts.

Without evidence of a causal link between the exclusive dealing provisions and Authenticom's foreclosure from the market, AutoLoop's claim fails. See El Aguila Food Prods., Inc. v. Gruma Corp., 301 F. Supp. 2d 612, 621 (S.D. Tex. 2003) (granting summary because "the evidence fails to establish that [the defendant's agreements] are the reason that some of the plaintiffs are suffering loss of shelf space with retailers"); Pitchford v. PEPI, Inc., 531 F.2d 92, 100 (3d Cir. 1975) (rejecting exclusive dealing claim where plaintiff "could not cite any examples of sales that were missed because of exclusive dealing practices by PEI during the damage period"). It is not enough for AutoLoop to show that Authenticom was cut out of the data integration market; it must also show that the exclusive dealing provisions, specifically, caused that foreclosure, at least in part.

AutoLoop insists that it is legally irrelevant whether CDK "might have caused the same damages" by lawful conduct. (See Authenticom Opp. at 72 (quoting Lee-Moore Oil Co. v. Union Oil Co. of Cal., 599 F.2d 1299, 1302 (4th Cir. 1979)).) But that does not move the needle. CDK is not entitled to summary judgment because its conduct might have been legal; it is entitled to summary judgment because AutoLoop has not brought forward evidence that CDK's vendor contracts, apart from its technological blocking, caused Authenticom's foreclosure from the market.

CDK makes essentially the same argument against the Dealers: because their alleged harm "does not flow from the alleged exclusive dealing provisions" but instead from what CDK calls its "unilateral decision[] to secure" its DMS, it has not suffered any injury. (CDK Dealers MSJ at 23.) The Dealers fail to respond to that argument, effectively conceding that summary judgment is warranted on their claim for injunctive relief.

2. Section 2 monopolization

Next, CDK moves for summary judgment on AutoLoop's § 2 monopolization claim. A firm violates the monopoly provision in § 2 when it: (1) possesses "monopoly power in the relevant market"; and (2) engages in "the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of superior product, business acumen, or historical accident." Viamedia, Inc. v. Comcast Corp., 951 F.3d 429, 451 (7th Cir. 2020) (quotation and citation omitted).

The court reserves ruling on this claim. The damages that AutoLoop claims flow from CDK's alleged monopolization are identical to the damages it seeks on its conspiracy claim. This court has previously held that CDK's unilateral decision to block third parties from accessing its DMS would be lawful but for its alleged conspiracy with Reynolds. See AutoLoop MTD Op., 362 F. Supp. 3d at 498. And, as discussed above, Defendants are entitled to summary judgment on AutoLoop's exclusive dealing claim because AutoLoop has

not presented evidence that the exclusive dealing provisions—independent of or in addition to technological blocking—foreclosed Authenticom from the integration market.

What remains is a § 2 monopolization claim premised solely on the alleged conspiracy. In other words, a reasonable jury could not award damages to AutoLoop for its monopolization claim unless the jury also finds that AutoLoop proved its conspiracy claim. But AutoLoop's damages model for those two claims are identical. (Israel Dep. [969-322] at 274:12-15 (Q: "You're saying that, if the conduct was unilateral, the damages are the same as your conspiratorial model, right?" A: "Correct.").) The court thus orders AutoLoop to show cause why the court should not dismiss AutoLoop's monopolization claim, as proving that claim to a jury would require proving the § 1 conspiracy yet would yield no additional damages. See Blue Cross & Blue Shield United of Wis. v. Marshfield Clinic, 152 F.3d 588, 593-94 (7th Cir. 1998) (discussing the issue of redundant violations).

D. State Law Claims

1. AutoLoop

AutoLoop also brings two Florida state antitrust claims (Counts IV and V), which rise and fall with its Sherman Act claims. See AutoLoop MTD Op., 362 F. Supp. 3d at 498 ("Federal and Florida antitrust laws are analyzed under the same rules and case law."); QSGI, Inc. v. IBM Glob. Fin., No. 11-80880-CIV, 2012 WL 1150402, at *4 (S.D. Fla. Mar. 14, 2012) (same for unfair practices claim). For the same reasons that CDK is not entitled to summary judgment on AutoLoop's § 1 conspiracy, it is also not entitled to summary judgment on AutoLoop's Florida claims based on that conspiracy. To the extent the Florida claims are based on exclusive dealing or monopolization, the court grants summary judgment for CDK on those claims.

2. Dealers

In addition to their federal claims, the Dealers brought several dozen claims under the consumer protection and antitrust laws of various states, most of which survived CDK's motion to dismiss. (See generally Dealers Compl. ¶¶ 209-736); see also Dealers MTD Op., 362 F. Supp. 3d at 548-57 (granting motion to dismiss counts VI, XXII, XXVII, XXXI, XXXIII, XXXVI, and XXXVIII). CDK now moves for summary judgment on the Dealers' remaining claims, which the court agrees largely track the federal claims. For the reasons discussed below, CDK's motion is granted in part and denied in part.

a. Global damages model

CDK's first argument applies to all the Dealers' damages claims. In short, CDK asserts that the Dealers' damages expert, Dr. Williams, calculated damages only for a "nationwide pass-through overcharge" class. (CDK Dealers MSJ at 15.) In other words, he calculated damages by assuming that costs from supracompetitive data integration prices were passed along by vendors to all affected Dealers nationwide. But, as CDK points out, Williams has not offered a way to "calculate indirect damages for the subset of states that recognize that form of relief." (Id.) Though Williams states in his report that he could calculate such damages using his proposed model, CDK argues that "the time to do so is long past." (Id. at 16.)

For their part, the Dealers respond that while they have not yet quantified damages by plaintiff or within each state, at summary judgment, they need only show that pass-through overcharges caused them damages. (See Dealers MSJ Opp. [1158] at 26; see also Dealers' Resp. to CDK's Supp. SJ Br. [1333] at 13 ("The allocation of state-by-state damages" for

states that allow indirect damages claims "is a matter of arithmetic and an extension of Dr. Williams's methodology.") The Dealers also submit that, if a nationwide class is certified under Illinois law, Williams's damages calculation would be accurate. (Id. at 28.) And if no nationwide class is certified, they urge, "state-by-state calculations can be made, if necessary, in connection with the class certification motion or at trial." (Id.) Calculating state-law damages at class certification is "customary," the Dealers submit, when antitrust cases involve states that (unlike under federal law) allow plaintiffs to recover indirect damages. (Dealers' Resp. to CDK's Supp. SJ Br. at 13.)

Though the court sympathizes with CDK's argument that it will be burdened by litigating these issues again at class certification, it denies summary judgment on damages. As anticipated by Judge St. Eve's initial case management order [166], class certification briefing—including expert reports related thereto—are set to take place after dispositive motions. At that stage, the Dealers will need to show more than just the generalized overcharge model they have presented so far. But that model is enough to establish that the Dealers suffered some damages, even if it does not quantity them by plaintiff or geography. Indeed, the court previously rejected CDK's Daubert motion aimed at alleged problems with Dr. Williams's methodology that were unrelated to Illinois Brick. See Daubert Op., 581 F. Supp. 3d at 1066-68.

The cases on which CDK relies are not clearly applicable. In one such case, Walton v. Trans Union, LLC, the court considered a motion for sanctions against a single plaintiff who provided "no specific information at all" regarding damages. See No. 1:07-cv-372, 2008 WL 2038598, at *1 (S.D. Ind. May 12, 2008). In another, the court found lost profits damages were barred by the "new business rule" as a matter of law. See Allergease, Inc. v. Walgreen Co., No. 15 C 4873, 2017 WL 66819, at *6 (N.D. Ill. Jan. 6, 2017). In a third, the plaintiff's expert admitted at his deposition that "he did not do any work to calculate damages." Dynegy Mktg. and Trade v. Multiut Corp., No. 02 C 7446, 2008 WL 2410425, at *8 (N.D. Ill. June 11, 2008).

Here, Dr. Williams provides a method to calculate damages, albeit no more than a nationwide figure. Williams will eventually need to calculate state-specific figures, but the court disagrees that Williams's methodology "does not even fit [the Dealers'] class theory." (CDK Dealers MSJ at 16.) As previously described, Williams in his report calculates an "Overcharge to Vendors" by a "Pass-Through Rate"—in other words, he estimates how much the vendors were overcharged by RCI and 3PA and what percentage was passed along to the Dealers. See Daubert Op., 581 F. Supp. 3d at 1062. That methodology is inconsistent with a direct damages class under federal law, but it fits well with the indirect damages classes the Dealers will seek to certify under state law.

In short, as Amerinet, Inc. v. Xerox Corp. (cited by CDK Dealers MSJ at 16) instructs, an antitrust plaintiff need not "prove exactly and with total certainty the amount of antitrust damages which it has sustained," but rather must provide a basis for the trier of fact to determine damages "without engaging in speculation or conjecture." 972 F.2d 1483, 1491-93 (8th Cir. 1992) (affirming summary judgment for defendant). In that case, unlike this one, the plaintiff's damages issues were closely tied to its causation issues: its own damages expert "strongly suggest[ed]" that the decline of plaintiff's business "was caused at least partly by, if not substantially or mainly" by factors other than the alleged antitrust violations. Id. at 1495. The court discussed quantification of damages

minimally and only in reasoning that the plaintiff's evidence on that score—speculation that, absent anticompetitive conduct, the plaintiff "would have sold" enough printers to stem its decline—was not enough to survive summary judgment. Id. at 1498. That is far different than the facts here, where Williams's methodology is sound enough for a jury, though it must be refined to provide greater geographical specificity. At first blush, the court agrees with the Dealers that such a determination is likely to be straightforward. If that is not the case, CDK will have the opportunity to demonstrate this at the appropriate time.

b. Massachusetts and New Jersey claims

Before addressing CDK's more detailed state-specific arguments, the court first grants summary judgment on Counts XLIV and XXXIX under New Jersey and Massachusetts law, respectively. Both of those states follow Illinois Brick and bar plaintiffs from seeking indirect damages. See In re Auto. Parts Antitrust Litig., No. 12-md-02311, 2013 WL 2456612, at *29 (E.D. Mich. June 6, 2013) (recovery under Massachusetts law "has not been extended to indirect purchasers"); In re Suboxone Antitrust Litig., 64 F. Supp. 3d 665, 703-04 (E.D. Pa. 2014) (New Jersey applies Illinois Brick to both its "antitrust law" and its "consumer protection law"); Sickles v. Cabot Corp., 379 N.J.Super. 100, 877 A.2d 267, 277 (2005) (same). Accordingly, the damages analysis discussed regarding the Dealers' federal conspiracy damages claim applies equally here and warrants summary judgment. See Section II.B.4.a, supra.

c. Nexus

CDK next argues that seven of the Dealers' claims fail because they lack nexus to the jurisdiction at issue. (CDK Dealers MSJ at 24-26.) The court has already granted judgment on the Dealers' Massachusetts claim (Count XXXIX), and the Dealers concede that their claim under District of Columbia law (Count IX) is deficient. (See Dealers MSJ Opp. at 29 n.41 (withdrawing claim).) The court denies summary judgment on the remaining five claims for the reasons below.

New Hampshire (Count XLIII): CDK argues that New Hampshire's consumer protection law applies only to "the conduct of any trade or commerce within" New Hampshire and the Dealers cannot show anything more than "increased prices paid in New Hampshire." (CDK Dealers MSJ at 25.) The court agrees that "numerous federal district courts" have acknowledged that New Hampshire law requires "proscribed conduct to occur within the state"—and not merely sales of overpriced goods or services; in the court's view, however, the Dealers have met that burden. See In re Pork Antitrust Litig., 495 F. Supp. 3d 753, 789 (D. Minn. 2020) (reviewing NHCPA caselaw). The Dealers cite evidence that CDK tracked and blocked 154 sets of login credentials, representing 17 New Hampshire dealerships. (See Nexus Facts Table [1158-2] at 6.) CDK does not address this allegedly anticompetitive conduct that arguably occurred within New Hampshire. That conduct distinguishes this case from those where the only nexus to New Hampshire takes the form of goods sold. Cf. In re Refrigerant Compressors Antitrust Litig., No. 2:09-md-02042, 2013 WL 1431756, at *17 (E.D. Mich. Apr. 9, 2013) (granting motion to dismiss where "there are no alleged conspirators taking action in New Hampshire").

Mississippi (Count XVII) : Mississippi law requires the Dealers to show "at least some conduct" by CDK that "was performed wholly intrastate." In re Microsoft Corp. Antitrust Litig., No. MDL 1332, 2003 WL 22070561, at *2 (D. Md. Aug. 22,

2003). CDK offers no analysis of that standard besides asserting that the Dealers cannot meet it. The Dealers submit that CDK blocked 87 sets of login credentials at 13 dealerships in Mississippi. (Nexus Facts Table at 7.) That blocking arguably took place within Mississippi, and the court finds that without a meaningful response from CDK, Dealers have met their burden at summary judgment.

New York (Count XXI): In support of its claim for summary judgment on the Dealers' New York claim, CDK cites caselaw establishing that federal antitrust laws preempt cases about conduct that "principally affects interstate commerce, with little or no impact on local or intrastate commerce." See H-Quotient, Inc. v. Knight Trading Grp., Inc., No. 03 Civ. 5889, 2005 WL 323750, at *4 (S.D.N.Y. Feb. 9, 2005) (dismissing claim). The Dealers cite several cases—to which CDK does not respond—suggesting that anticompetitive conduct may be challenged under New York law even if it does not primarily affect intrastate commerce, and that New York accepts price inflation as evidence of impact. (See Dealers MSJ Opp. at 36 & n.52.) The Dealers have presented evidence of substantial blocking and supracompetitive prices within New York. (See Nexus Facts Table at 9-10.) Summary judgment on the New York claims is denied.

South Dakota (Count XXVI): South Dakota law merely requires that the Dealers show "an unlawful conspiracy to restrain trade or commerce, any part of which is within the state of South Dakota." In re Dynamic Access Memory Antitrust Litig., 516 F. Supp. 2d 1072, 1099 (N.D. Cal. 2007) (granting motion to dismiss where complaint alleged only "a substantial effect on the foreign and interstate commerce of the United States"). Because the Dealers show allegedly conspiratorial activity within South Dakota (see Nexus Facts Table at 12-13), the court denies summary judgment.

West Virginia (Count XXX): CDK correctly (and unremarkably) asserts that "West Virginia's antitrust law is directed towards intrastate commerce." See State ex rel. Palumbo v. Graley's Body Shop, Inc., 188 W.Va. 501, 425 S.E.2d 177, 183 n.11 (1992). But, like South Dakota, West Virginia's Antitrust Act targets anticompetitive behavior "any part of which" is within the state. (See Dealers MSJ Opp. at 38 (discussing W. Va. Code § 47-18-4).) For the same reasons discussed above, the court denies summary judgment. (See Nexus Facts Table at 14 (discussing blocking efforts in West Virginia).)

d. Antitrust claims under consumer protection laws

CDK next argues that five counts must be dismissed because the statutes on which they are based—prohibiting "deceptive," "false," or "unconscionable" conduct—do not reach the kind of anticompetitive conduct alleged by the Dealers. (See CDK Dealers MSJ at 26 & n.8.) To those five claims, CDK devotes one brief paragraph and one footnote with one case citation for each state, some of which lack even a parenthetical explanation of their relevance. The court is tempted to ignore CDK's underdeveloped "litigation by footnote" entirely. But to streamline the case for trial, the court opts to consider the issues.

CDK's argument relies on the court's Rule 12(b)(6) ruling, in which it dismissed the Dealers' claim under Arkansas's Deceptive Trade Practices Act because the Dealers failed to allege "fraudulent and/or unconscionable acts" as opposed to "'unfair' businesses practices." Dealers MTD Op., 362 F. Supp. 3d at 556. The Dealers respond that Colorado's and New Mexico's laws in fact do prohibit "unfair" conduct. (See Dealers MSJ Opp. at 39-40 & nn. 56-57); see also In re Remicade Antitrust

Litig., 345 F. Supp. 3d 566, 589 (E.D. Pa. 2018) (allegation that conspiracy caused inflated consumer prices established unfairness or unconscionable conduct under New Mexico law). The court therefore denies summary judgment on those counts.

The court grants summary judgment on the other three counts—those under Minnesota law (Count XL), North Dakota law (Count XLVI), and South Dakota law (XLVIII). Though the court accepts that courts in those states interpret their consumer protection laws broadly, the case law presented by CDK is persuasive. For example, in In Re New Motor Vehicles Canadian Export Antitrust Litigation—which involved an alleged conspiracy to block imported Canadian vehicles and keep new vehicle prices in the United States artificially high—the court held that the plaintiffs had not shown fraud or deception. 350 F. Supp. 2d 160, 176 (D. Me. 2004). In so holding, the court rejected the plaintiffs' assertion that the defendants deceived car buyers when they "fail[ed] to inform consumers of the conspiracy to keep out Canadian cars." Id. The court reasoned that the anticompetitive conspiracy is what hurt car buyers, not deceptive marketing surrounding it. See id. at 176-77 ("[E]ven if an American dealer had said to a consumer, 'You know, this car costs more because we have conspired with the manufacturers to prevent Canadian cars from coming into the American market,' the price would have remained the same (unless and until someone halted the alleged conspiracy)."). The court therefore dismissed the plaintiffs' claims under South Dakota and North Dakota law. See id. at 197-98, 202-03.

The Dealers respond that they have indeed "allege[d] fraudulent and deceptive practices," but do not explain further, nor do they cite cases from either state in which an antitrust conspiracy like this one was characterized as fraudulent. (See Dealers MSJ Opp. at 39 n.58.) Dealers discuss one antitrust case where a South Dakota consumer protection claim survived, but that case involved specific "misrepresentations" that were made to government bodies and were allegedly fundamental to the conspiracy. See In re DDAVP Indirect Purchaser Antitrust Litig., 903 F. Supp. 2d 198, 229 (S.D.N.Y. 2012).

Finally, the Dealers argue that actions undertaken to "maintain [the defendant's] monopoly on the market" can sustain a claim under Minnesota's consumer protection statute. (Dealers MSJ Opp. at 40.) But in the case on which the Dealers rely, the plaintiffs' claim was based on the defendant's "baseless legal argument in a legal proceeding[] and to the Patent and Trademark Office," both done "in order to maintain its monopoly." Sheet Metal Workers Local 441 v. GlaxoSmithKline, PLC, 737 F. Supp. 2d 380, 414 (E.D. Pa. 2010). The court's reasoning that a baseless argument made to maintain a monopoly could represent a "misleading statement or deceptive practice" made "in connection with the sale of any merchandise" does not mean that general conspiratorial conduct is a "misleading statement or deceptive practice." Id. Here again, the court finds New Motor Vehicles more persuasive. In that case, the court dismissed the Minnesota claim, citing to a parallel action in Minnesota state court in which the judge rejected the plaintiffs' argument that "it was deceptive to state a price for a car while knowing that the conspiracy maintained prices at an artificially high level." 350 F. Supp. 2d at 190. That is what the Dealers would have to argue here. Summary judgment is granted on this claim.

e. Standing

CDK next argues that the Dealers are not proper plaintiffs under Nevada and West Virginia law because each restricts standing to specific classes to which the

Dealers do not belong. The court denies CDK's motion as to both states.

Nevada (Count XLII): According to CDK, Nevada's Deceptive Trade Practices Act only permits suits "by an elderly person or a person with a disability." (See CDK Dealers MSJ at 27 (quoting In re Wellbutrin XL Antitrust Litig., 260 F.R.D. 143, 163 (E.D. Pa. 2009) (citing Nev. Rev. Stat. § 598.0977)).) But Nevada's law—as its broad name suggests—is in fact not so limited. To the contrary, "any person who is a victim of consumer fraud" may bring an action under the statute. See DDAVP Antitrust Litig., 903 F. Supp. 2d at 226-27 (citing Nev. Rev. Stat. § 41.600, which defines a "victim of consumer fraud" as someone subject to a "deceptive trade practice" as defined in Nev. Rev. Stat. § 598.0915 to 598.0925). The court acknowledges a split in the caselaw on this score, but the cases cited by CDK limiting the statute's scope make no mention of § 41.600, which provides a cause of action for "any person" who is a victim of a deceptive trade practice. (See CDK Dealers Reply [1123] at 13-14.) Nor does CDK present a different reading of § 41.600; it merely notes that § 598.0977 is "almost entirely superfluous" if § 41.600 provides a right of action for non-elderly or disabled plaintiffs. (Id.) The court disagrees: § 598.0977 gives elderly or disabled plaintiffs collection priority and provides for punitive damages.²³ See Vrugtman v. It's Just Lunch Int'l LLC, No. EDCV 20-2352, 2021 WL 4979443, at *6 (C.D. Cal. Sept. 24, 2021) (noting that the NDTPA "only authorizes punitive damages for a deceptive trade practice" for elderly or disabled plaintiffs and that § 41.600(2)—the "general provision" for deceptive practice claims—does not permit recovering punitive damages). The allowance of priority status and award of punitive damages for the elderly does not eliminate the right of all other plaintiffs to recover damages for consumer fraud. Summary judgment is denied.

23 The court agrees with CDK, however, that the court in In re DDAVP wrongly concluded that elderly or disabled plaintiffs could "recover more for each violation" than others, suggesting that such plaintiffs are entitled to recover "civil penalties" under § 598.0973. See 903 F. Supp. 2d at 227. As CDK correctly notes, § 598.0973 permits collection of "civil penalties," only by "certain qualified public officials." See Tretiak v. Option One Mortg. Corp., 381 P.3d 670 (Table) (Nev. May. 11, 2012).

West Virginia (Count XLIX) : CDK urges that West Virginia's Consumer Credit and Protection Act applies only to consumers who are "natural person[s]." (See CDK Dealers MSJ at 27.) The plain language of the statute dooms that position. West Virginia Code § 47A-6-106 creates a cause of action for "any person" and § 46A-1-102 defines "person" to include an "organization," which is in turn defined to include "a corporation." While CDK responds that "the West Virginia Supreme Court has repeatedly emphasized the consumer-oriented nature of the statute" (CDK Dealers Reply at 14), the court declines to override the plain language of the statute and denies summary judgment.

f. State class action bars

CDK moves for summary judgment on three putative class claims under Illinois (Count XI), Colorado (Count XXXV), and South Carolina (Count XLVII) law. (CDK Dealers MSJ at 27-32.) According to CDK, each of those states' laws prohibit class actions for indirect damages. And CDK asserts that those state statutes apply to state law claims brought in federal court, too, notwithstanding the Supreme Court's ruling in Shady Grove Orthopedic Associates., P.A. v. Allstate Insurance Co., 559 U.S. 393, 130 S.Ct. 1431, 176 L.Ed.2d 311 (2010). CDK's motion is denied. i. Shady Grove

The parties dispute whether the plurality opinion or Justice Stevens' concurring opinion in Shady Grove applies. In that case, the Supreme Court considered whether a federal class action under Rule 23 was barred by a New York state law prohibiting class actions in suits seeking penalties or statutory damages. 559 U.S. at 396, 130 S.Ct. 1431. The plaintiff, a medical office, sued one of its patient's insurers for failing to pay on time and then refusing to pay statutory interest that had accrued on the overdue benefits. Id. at 397, 130 S.Ct. 1431. The district court dismissed the suit on jurisdictional grounds, reasoning that statutory interest was a "penalty" under New York law and could not be maintained as a class action, meaning that the plaintiff's roughly $500 claim fell short of the amount-in-controversy requirement for diversity jurisdiction. Id. The Second Circuit affirmed. Id. at 398, 130 S.Ct. 1431.

The Supreme Court reversed. Writing for a four-Justice plurality, Justice Scalia concluded that New York's prohibition conflicted with Rule 23, which trumped state law "regardless of its incidental effect upon state-created rights." Id. at 410, 130 S.Ct. 1431. Justices Steven Concurred in the judgment, but cautioned that Rule 23 could not "displace a state law that is procedural in the ordinary use of the term but is so intertwined with a state right or remedy that it functions to define the scope of the state-created right." Id. at 423, 130 S.Ct. 1431 (Stevens, J., concurring). Recognizing that the analysis is often complex, Justice Stevens reasoned that the New York law at issue was not "so intertwined" as to affect substantive rights because the statute applied "not only to claims based on New York law but also to claims based on federal law or the law of any other State." Id. at 432, 130 S.Ct. 1481. In other words, class actions for statutory damages cannot be brought in New York state court "under any source of law." Id. at 436, 130 S.Ct. 1431.

CDK argues that "[b]ecause no single opinion in Shady Grove garnered a majority, this court must apply the principle of Marks v. United States, 430 U.S. 188, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977)" to determine which opinion—Scalia's plurality or Stevens' concurrence—controls. (CDK Dealers MSJ at 28-29.) Marks, in turn, instructs that this court must apply the "position taken by those Members who concurred in the judgments on the narrowest grounds." 430 U.S. at 193, 97 S.Ct. 990. According to CDK, that position would have to be Justice Stevens' opinion, whereas the Dealers respond that Justice Stevens' opinion "is not logically narrower." (Dealers MSJ Opp. at 44.) As even the cases cited by the Dealers recognize, "most" courts to consider this question "have concluded that the Stevens concurrence controls." See In re Aggrenox Antitrust Litig., No. 3:14-md-2516, 2016 WL 4204478, at *5 (D. Conn. Aug. 9, 2016).

The court agrees with that majority. Absent additional guidance from the Seventh Circuit,²⁴ the court concludes that

24 In the court's motion to dismiss opinion, it declined to determine "which standard" applied under Shady Grove because CDK had not "identified any authority for concluding that the class action bars at issue are so intertwined with a state-created right or remedy as to justify finding that it trumps Rule 23" even "under Justice Stevens's analysis." Dealers MTD Op., 362 F. Supp. 3d at 553 & n.23. The court noted that the Seventh Circuit had "not squarely addressed" the Shady Grove issue but "otherwise has indicated that Justice Scalia's plurality sets forth the controlling legal standard." Id.; see Sawyer v. Atlas Heating & Sheet Metal Works, Inc., 642 F.3d 560, 564 (7th Cir. 2011) (Shady Grove "holds that Rule 23 applies to all federal civil suits, even if that prevents achieving some other objective that a court thinks valuable"); State Farm Life Ins. Co. v. Jonas, 775 F.3d 867, 869 (7th Cir. 2014) ("[F]ederal procedures govern in federal litigation") (citing Shady Grove). With further thought, however, the court agrees with CDK that neither opinion from the Seventh Circuit considered the "plurality or concurrence" question or addressed a state class action bar arguably preempted by Rule 23. (See CDK Dealers MSJ at 29-30.)

Justice Stevens' opinion is indeed narrower. Both opinions recognize that the Federal Rules cannot "abridge, enlarge or modify any substantive right." 28 U.S.C § 2072(b); see Shady Grove, 559 U.S. at 410, 130 S.Ct. 1431 ("We understand [the concurrence] to accept the framework we apply"); id. at 424 n.8, 130 S.Ct. 1431 (Stevens J., concurring) (noting that the plurality agrees that "courts must look to whether a federal rule alters substantive rights in a given case"). They disagree, however, about how to understand when a rule alters a "substantive right." In the plurality's view, courts need only look to the federal rule itself, and "if it governs only the manner and the means by which the litigants' rights are enforced," it does not alter substantive rights. Id. at 407, 130 S.Ct. 1431. But Justice Stevens' concurrence requires courts to take another step and look to whether the state law that Rule 23 would preempt is "so bound up with the state-created right or remedy that it defines the scope of that substantive right or remedy." Id. at 420, 130 S.Ct. 1431. In other words, Justice Stevens' view is that the plurality's requirement—that the federal rule must be procedural—is necessary, but not sufficient. Justice Stevens' analysis is therefore "narrower."²⁵

25 The Dealers do not offer any case where a court has reasoned through Shady Grove and held that the plurality opinion controls. At most, Aggrenox concludes that Justice Stevens' concurrence is narrower "only in the sense that it would reject state procedural rules in fewer cases," but is "not logically narrower" because the Justices in the plurality "do not implicitly approve of its rationale for sometimes allowing state procedural rules to control" and in fact "explicitly reject that rationale." 2016 WL 4204478 at *5. That court called the concurrence "Marks-doctrine dicta rather than [a] Marks-doctrine holding." Id. at *6. But it also recognized that "the portions of Scalia's opinion with which Stevens did not agree did not garner a majority" and could not therefore control. Id. "So it may be that in a case where a state procedural rule is part of the state's framework of substantive rights, there simply is no controlling Supreme Court holding and that Marks-doctrine dicta is the most persuasive guide." Id. In other words, the Aggrenox followed the roadmap the court will follow here.

ii. As Applied

The Illinois Antitrust Act ("IAA") authorizes indirect purchasers to sue but provides that "no person shall be authorized to maintain a class action in any court of this State for indirect purchasers asserting claims under this Act, with the sole exception of this State's Attorney General, who may maintain an action parens patriae." 740 ILCS § 10/7(2). As to whether the Act "functions to define the scope of the state-created right," Shady Grove, 559 U.S. at 423, 130 S.Ct. 1431, the parties have not identified any binding Seventh Circuit authority and instead cite conflicting district court opinions, both from within this district and outside of it.

Given the conflicting caselaw, the court returns to the analysis in Shady Grove. In his concurrence, Justice Stevens noted that "the bar for finding an Enabling Act problem is a high one," meaning that "[t]he mere possibility that a federal rule would alter a state-created right is not sufficient." Id. at 432, 130 S.Ct. 1431. He emphasized that the text of New York's law "expressly and unambiguously applies not only to claims based on New York law but also to claims based on federal law or the law of any other State." Id. Accordingly, he found it "hard to see how" it "could

be understood as a rule that, though procedural in form, serves the function of defining New York's rights or remedies." Id.

The IAA is, in that respect, different than the New York rule at issue in Shady Grove. Unlike the New York rule—which applied to every claim for statutory damages or penalties under any law—the prohibition against class actions in the IAA is specific to "indirect purchasers asserting claims under this Act." 740 ILCS § 10/7(2). CDK believes that distinction is crucial. And many courts have indeed "distinguished between 'pan-statutory' class-action bars," like the one at issue in Shady Grove, and "limitations built into particular state statutes," like the one at issue here. See Delgado v. Ocwen Loan Servicing, LLC, No. 13-CV-4427, 2017 WL 5201079, at *9 (E.D.N.Y. Nov. 9, 2017) (citing cases). Under that logic, many courts—perhaps a majority—have concluded that Rule 23 must give way to the IAA's class action bar. See, e.g., In re Wellbutrin XL Antitrust Litig., 756 F. Supp. 2d 670, 677 (E.D. Pa. 2010) ("Because the indirect purchaser restrictions of the IAA are 'intertwined' with the underlying substantive right, application of Rule 23 would 'abridge, enlarge or modify' Illinois' substantive rights, and therefore Illinois' restrictions on indirect purchaser actions must be applied in federal court."); In re Opana ER Antitrust Litig., 162 F. Supp. 3d 704, 723 (N.D. Ill. 2016) (same).

But, in the ways that are most relevant, the IAA is analogous to the rule at issue in Shady Grove. Unlike the Illinois Brick bar, which is "certainly substantive in that it affects the 'rights and remedies' of indirect purchasers," the IAA's restriction on class actions "does not affect their substantive rights." In re Broiler Chicken Antitrust Litig., 290 F. Supp. 3d 772, 818 (N.D. Ill. 2017). As the Aggrenox court thoughtfully reasoned, it is difficult to "square Shady Grove's allowance of a Rule 23 class action despite New York's class-action bar with the dis allowance of a Rule 23 class action in the case of Illinois's class-action bar simply on the basis that Illinois's bar is narrower." 2016 WL 4204478, at *6. True, it is narrower in that "its application is limited to indirect-purchaser antitrust claims" and it is tied "more specifically to particular substantive rights." Id. "[B]ut if New York's state-law bar is not a procedural rule that alters the scope of a substantive right or remedy, then the narrower scope of Illinois's state-law bar does not make it one that does." Id. Of course, in practice, the ability to bring a class action may significantly increase litigation and "could have a profound effect" on a defendant's "total liability." Id. As Justice Stevens recognized, "any device that makes litigation easier makes it easier for plaintiffs to recover damages." Shady Grove, 559 U.S. at 434, 130 S.Ct. 1431 (Stevens, J., concurring). But Shady Grove instructs that a procedural rule is not rendered substantive solely by its effect on damages. Cf. id. at 436, 130 S.Ct. 1431 (Ginsburg, J., dissenting) (criticizing the majority's ruling for "transform[ing] a $500 case"—the statutory penalty available to a single plaintiff—"into a $5,000,000 award" for a class). Summary judgment is denied.

The South Carolina and Colorado statutes are effectively identical, and the court denies summary judgment on those claims, too. The South Carolina Unfair Trade Practices Act ("SCUTPA") authorizes suit by an aggrieved party "individually, but not in a representative capacity, to recover actual damages." S.C. Code § 39-5-140. CDK cites cases dismissing SCUTPA claims under Shady Grove, but those cases are based on the same logic the court rejected above. See, e.g., In re TD Bank, N.A., 150 F. Supp. 3d 593, 635 (D.S.C. 2015) (holding without

explanation that "SCUTPA is importantly different from the state procedural law at issue in Shady Grove, because the New York law had no substantive component" whereas "the prohibitions against class actions ingrained in the very text of the SCUTPA ... are substantive portions of South Carolina law"). The same analysis applies to the Colorado Consumer Protection Act. See Col. Rev. Stat. § 6-113(2) (permitting suits for deceptive trade practices "[e]xcept in a class action"). In short, the state class actions bars cited by CDK are preempted by Rule 23 and do not apply here.

g. Statutory Notice Requirements

Finally, CDK moves for summary judgment on five claims because the Dealers failed to give proper notice either to the state attorney general (Count VII (Arizona), Count X (Hawaii), Count XXVII (Utah), and Count XLIV (New Jersey)) or to CDK itself (West Virginia (Count XLIX)). Two of those claims need not be addressed: the court has already granted summary judgment on the New Jersey claim (see Section II.D.2.b, supra), and CDK now concedes the Dealers met Arizona's notice requirement (CDK Dealers Reply at 19). That leaves the claims arising under Hawaii, Utah, and New Jersey law. The court denies summary judgment as to those claims.

In Hawaii, indirect purchasers seeking to bring a class action for antitrust violations must serve a copy of their complaint "not later than seven days after filing," and the attorney general thereafter has sixty days to decide whether to take up the case. Haw. Rev. Stat. § 480-13.3(a)(1), (5). Utah's code states that the "attorney general shall be notified by the plaintiff about the filing of any class action involving antitrust violations that includes plaintiffs from this state." Utah Code § 76-10-3109(9). West Virginia law in effect at the time of the Dealers' suit²⁶ requires a plaintiff to "inform[] the seller" in writing "of the alleged violation" and provide time "to make a cure offer": twenty days from receipt or "ten days in the case a cause of action has already been filed." See W. Va. Code § 46A-6-106(c) (West 2015).

26 In 2021, after this suit was filed, the statute's cure requirement was moved to a different subsection. See W. Va. Code § 46-A-6-106(c) (cure requirement removed); Id. § 46-A-5-108 (requiring notice and opportunity to sue under "§ 46A-6-1 et seq. of this code").

The court must first engage in a Shady Grove analysis. If the court finds that Federal Rule of Civil Procedure 23 conflicts with various state notice requirements, those requirements would be displaced and unenforceable in federal court. (Even more so than the class-action bars discussed above, the notice requirements are procedural—not substantive—requirements.) If, on the other hand, "there is no conflict" between Rule 23 and the notice requirements, "then the state notice requirements would apply," as this court previously reasoned. See Dealers MTD Op., 362 F. Supp. 3d at 555 (declining to rule on this issue because it was insufficiently briefed). As with the class-action bars, the parties agree there is no binding Seventh Circuit law, and each side presents district court opinions in their favor.

CDK, citing several such cases, urges that Shady Grove does not apply and the notice provisions therefore bar the Dealers' claims. (CDK Dealers MSJ at 33-34.) Rule 23 "is not sufficiently broad to cover the state statutory notice provisions," CDK claims, because those provisions "do not attempt to answer the same question or subject." In re Lipitor Antitrust Litig., 336 F. Supp. 3d 395, 415 (D.N.J. 2018); see also In re Loestrin 24 FE Antitrust Litig., 410 F. Supp. 3d 352, 375 (D.R.I. 2019) (same); In re Asacol

Antitrust Litig., No. 15-cv-12730, 2016 WL 4083333, at *14-15 (D. Mass. July 20, 2016).

The Dealers cite other cases, though most move straight to deciding whether and how to apply the Scalia plurality or Stevens concurrence without first deciding whether the notice requirements in fact conflict with Rule 23. See, e.g., In re Aggrenox Antitrust Litig., 94 F. Supp. 3d 224, 254 (D. Conn. 2015) (reasoning that Hawaii's notice requirements are not "sufficiently a 'part of a State's framework of substantive rights or remedies' to be controlling in federal court even under the Stevens concurrence"). But the court is persuaded by the reasoning in In re Restasis Antitrust Litigation, which did find that "Hawaii's [notice] law conflicts with Rule 23." See 355 F. Supp. 3d 145, 155-56 (E.D.N.Y. 2018). The court in that case accurately observed that "a state law that restricts the types of claims eligible for class treatment beyond the limits established by Rule 23 conflicts with the federal rule." Id. at 156 (citing Shady Grove). The court then held: "The situation here is identical. Hawaii's law includes a limitation on the types of claims eligible for class treatment and this limitation is absent from Rule 23," meaning that it conflicts and is preempted under Shady Grove. Id.

Though the court is sympathetic to CDK's logic—a statute forbidding class actions does feel different than a notice requirement to bring a class action—the Supreme Court was not. CDK's position is akin to the one adopted by the Second Circuit in Shady Grove, a position which the Supreme Court reversed. As Justice Scalia's plurality decision summarizes it: "[t]he Second Circuit believed that [New York's class-action bar] and Rule 23 do not conflict because they address different issues." Shady Grove, 559 U.S. at 399, 130 S.Ct. 1431. "Rule 23, it said, concerns only the criteria for determining whether a given class can and should be certified" whereas New York's law "addresses an antecedent question: whether the particular type of claim is eligible for class treatment in the first place—a question on which Rule 23 is silent." Id. That is almost exactly how CDK describes the notice requirements at issue here: as a "condition precedent for class treatment," which Rule 23 "does not address ... one way or the other." (CDK Dealers Reply at 18.)

As previewed above, the court holds that—even under Justice Stevens' test, which preserves state procedural rules that are "part of a State's framework of substantive rights or remedies," 559 U.S. at 419, 180 S.Ct. 1431—Rule 23 preempts the notice statutes. If, as the court held above, a prohibition on class actions entirely is not part of that substantive framework, then the requirement of notice to the attorney general is also not part of that framework. Though CDK half-heartedly insists that the notice requirements are enforceable under Shady Grove, it identifies no authority for that position. Instead, CDK cites one case that found Shady Grove did not apply and two cases dealing with a completely different kind of notice—notice to the defendant (required by a state rule or by caselaw) that its actions were illegal. (CDK Dealers MSJ at 34.)

* * *

For the reasons detailed above, the court grants summary judgments on Counts IX, XLIV, XXXIX, XL, XLVI, and XLVIII. CDK's motion for summary judgment is otherwise denied.

III. Reynolds's Motion for Summary Judgment on MVSC's Claims

Reynolds moves for summary judgment on MVSC's remaining claims for conspiracy in violation of Sherman Act §§ 1 and 2 and related California and Illinois state

laws [954]. In short, MVSC claims that: (1) Reynolds conspired with CDK both to block independent data integrators—based on the same theories discussed at length above; and (2) Reynolds and CDK conspired to deny MVSC access to "certified" integration on their respective DMSs to benefit their joint venture—Computerized Vehicle Registration ("CVR")—which was one of MVSC's main competitors. MVSC settled with CDK and CVR in October 2019 [778], but its claim against Reynolds survives.

At this stage, Reynolds argues—in addition to the points addressed above with respect to the conspiracy generally—that MVSC has not marshaled sufficient evidence to show that Reynolds conspired with CDK to specifically block MVSC's access to data integration. Reynolds also argues that the economic terms it offered MVSC for membership in RCI were non-discriminatory and did not reflect an effective refusal to deal.

A. Background

The following facts are taken primarily from the parties' Local Rule 56.1 statements. These facts are undisputed unless otherwise noted.

1. The parties

CVR—a joint venture between CDK and Reynolds since 1992—provides electronic vehicle registration ("EVR") and titling services in 18 states. Reynolds owns 20 percent of CVR; CDK owns the rest. (See MVSC's Responses to Reynolds's Statement of Undisputed Material Facts [1079] ("MVSC SOFR") ¶ 3.)

Beginning in the 2000s, new providers with web-based EVR solutions (which CVR at the time did not offer²⁷) entered some states in which CVR already operated. (See DRPSOF ¶ 97.) MVSC was one of these new competitors. (Id.) MVSC provides EVR services in four states: California (where it has the exclusive endorsement of the state's dealer association); Oregon (where it is the only EVR provider); Virginia; and Illinois. (MVSC SOFR ¶ 1.) CVR suffered significant loss of market share to MVSC in 2013 and 2014. (DRPSOF ¶ 99.)

27 According to CVR's corporate representative, CVR offered a "client-based platform," meaning that "each user has the software installed on their computer" unlike a "web-based program" where the software is "in the cloud." (See Medina 30(b)(6) Dep. [1069-53] at 147:14-24.) According to an internal CVR document describing the "primary competitive disadvantages CVR faces" with internet-based competitors, data in the old CVR system was "distributed on separate Delphi PCs, until such time that title clerks 'submit' the deal to our back end processing system." (See [1069-126] at 8.)

2. EVR regulations

EVR providers partner with state departments of motor vehicles ("DMVs") to provide services to dealers. EVR providers must receive state approval before beginning operations. Regulations vary by state. Some, like California and Illinois, cap (or effectively cap) the fees that EVR providers may charge. For example, Illinois car dealers cannot charge more than $25 per EVR transaction; of that $25, up to $10 goes to the EVR provider and the rest goes to the dealer. In California, EVR providers charge dealers $30 per transaction, $5 of which goes to the DMV and $25 of which goes to the provider. (See DRPSOF ¶ 92.)

To perform their services, EVR providers need three categories of data: (1) car buyer identification; (2) vehicle identification; and (3) financing information. (MVSC SOFR ¶ 9.) That data can be accessed from a dealer's DMS in one of three ways: (1) through certified integration from the DMS provider; (2) through manual reporting tools built into the DMS that allow

dealers to export and transmit reports to third party apps of their choosing; or (3) by independent third-party integrators like Authenticom. As discussed below, MVSC disputes that the second option—manual reporting—provides functionality equivalent to certified or third-party integration. (Id. ¶ 10.)

Each state has its own EVR approval process, fees, and regulations around EVR services, meaning that a provider approved to provide EVR services in one state may or may not be approved to operate in another. (Id. ¶ 8.) Certain "point-of-sale" states—including California (as of January 1, 2019), Illinois, and Virginia—require EVR transactions to be initiated during the sale of the vehicle at the dealership. (Id. ¶ 6.) To satisfy that requirement, MVSC maintains that it needs "near real-time" access to data on the DMS, which, until recently, could only be provided through certified access programs (like RCI and 3PA) or independent integrators. (See DRPSOF ¶ 93.) Reynolds disputes that "near real-time access" is required to initiate EVR process at the point of sale. (Id.) Reynolds points to: (1) MVSC COO Joe Nemelka's testimony that an EVR provider can be competitive in point-of-sale states using manual reporting tools if the difference is one of seconds; and (2) an MVSC email stating that manual reporting takes "about 30 seconds." (Id.) Reynolds points out that MVSC has manually processed EVR transactions in other point-of-sale states, as have other EVR providers. (See id.)

3. MVSC's EVR service

It is undisputed that during the relevant period, MVSC was able to use manual reporting tools to process EVR transactions in non-point-of-sale states. (MVSC SOFR ¶ 11.) Further, for much of the alleged conspiracy period, manual tools were the near-exclusive method MVSC used to obtain data from dealers on a Reynolds DMS. (Id. ¶ 13.) MVSC assured its dealer customers that "we do not need to participate in the 3PA program to securely get the data from your DMS and provide you all the features and services we currently provide," and explained that dealers "can simply run a manual report and deliver us the data" which "takes about 30 seconds every morning." (Id. ¶ 16.) MVSC admits it made such marketing claims, but nevertheless disputes that manual reporting options are truly equivalent to automated integration or are in fact sufficient for MVSC to provide innovative services. (Id. ¶¶ 11, 16.) It also disputes that all dealerships were willing to manually export their data. (Id. ¶ 16.) MVSC argues, further, that Reynolds made the manual reporting process more difficult by disrupting or limiting dealerships' ability to run such reports. (See id. ¶ 14.) Reynolds insists, to the contrary, that it has never cut off or blocked dealers from using manual reporting tools to export and transmit data to MVSC.

In late 2017, stymied in its efforts to obtain real-time access to DMS data, MVSC developed a software tool called "Electron" that works in conjunction with Reynolds's manual reporting tools to access data at near real-time speeds. (Id. ¶ 18.) According to Nemelka, MVSC developed Electron because manual reporting did not provide MVSC with real-time data access sufficient to process EVR transactions in point-of-sale states. (DRPSOF ¶ 94.) Reynolds claims MVSC spent no more than a "couple hundred thousand dollars" to develop Electron (MVSC SOFR ¶ 18), and that Electron merely "augment[ed] preexisting free manual reporting tools provided by the DMS providers" and "save[d] dealers a few seconds or clicks when they use manual reporting." (DRPSOF ¶ 119.) MVSC responds that the figure cited by Reynolds reflects only the amounts paid to software developers and does not include millions of dollars' worth

of MVSC's own employees' labor devoted to the project. (See MVSC SOFR ¶ 18.) Klein (MVSC's expert) quantified workaround costs "conservatively" at $4.895 million. (See DRPSOF ¶ 119.)

MVSC admits that the Electron program enables it to provide EVR services in point-of-sale states and that Defendants have not blocked that use. (MVSC SOFR ¶ 19.) However, MVSC cites evidence—including Klein's report and testimony and Nemelka's declaration—that some customers who would otherwise purchase EVR services from MVSC do not consider Electron to be equivalent to automated integration because Electron is manual and therefore slower and more error-prone than automated integration. (Id.; see also DRPSOF ¶ 111)

4. CVR's EVR service

CVR generates approximately $100 million in annual revenue, mostly from Illinois (where CVR is the leading EVR provider) and from California. Reynolds, as part owner of CVR, receives 20 percent of its dividend distributions. (DRPSOF ¶ 95.) Reynolds also receives 35 percent of the transaction revenues for CVR transactions processed on the Reynolds DMS, which it characterizes as "royalties" rather than integration fees. (See MVSC SOFR ¶ 22.) In 2014, Reynolds received $13.4 million in dividends and $6.97 million in royalties. (DRPSOF ¶ 95.)

According to Reynolds, CVR relied exclusively on a process "akin to manual reporting" to access Reynolds's own DMS data up through 2016 and to this day uses manual reporting in most states where it does not have a web-based EVR app. (MVSC SOFR ¶ 21.) MVSC—whose liability theory depends in part on being denied access to automated, certified integration—disputes that CVR's historical method of integration was "akin" to the kind of manual reporting MVSC used and notes that Reynolds, as part-owner of CVR, is obligated to ensure that CVR has a functional, guaranteed means of data access in all markets, including point-of-sale states. (Id.)

5. MVSC's application to CDK's 3PA program

MVSC applied to join CDK's 3PA program five separate times. CDK denied MVSC's February and July 2014 applications. (DRPSOF ¶ 105.) Internal emails between CDK executives reflect that CDK declined to allow MVSC to use 3PA for certified integration because MVSC was CVR's "biggest competitor in California, and the dealers they've sold are really big ones — costing [CVR] in the millions in revenue," which made certifying MVSC for 3PA access "too painful to consider." (See [1089-5]; see also [1069-134] (internal CDK email where Dan McCray says: "Recall we turned [MVSC] down as they were a direct competitor to CVR."); McCray Dep. [1069-27] at 154:20-24 (McCray testifying that CDK refused to allow MVSC to join 3PA "because MVSC was a competitor" to CVR).) In response to MVSC's third application, CDK proposed that MVSC return a monthly fee of 35 percent of revenue; in response to the fourth and fifth application, CDK proposed a 25 percent fee. MVSC argues these fees were unreasonable and economically infeasible for MVSC to pay. (Id. at ¶¶ 106-07.) Nemelka testified that MVSC projected CDK's proposed terms would have cost it "tens of millions" annually. (MSVC SOFR ¶ 52.)

6. MVSC's application to Reynolds's RCI program

Unlike CDK, Reynolds did not enforce "category restrictions" on its customers—in other words, it was theoretically willing to do business with EVR providers other than CVR.²⁸ But MVSC argues that the

28 Reynolds cites a 2014 CDK PowerPoint noting that Reynolds had no "category restrictions" on admitting CVR's competitors into RCI: "No Categories/Category Restrictions (ex: CVR competitors included)[.]" (MVSC SOFR ¶ 53.) MVSC is skeptical; it cites internal emails from CDK, CVR, and Reynolds (discussed in detail below) in which Scott Herbers—the CDK executive overseeing the CVR joint venture—wrote that salespersons could tell prospective clients that MVSC "does not and WILL NOT have direct DMS integration" with Reynolds and said that MVSC "will NEVER have access" to data through Reynolds. (See 1069-131].)

terms Reynolds proposed when MVSC applied for RCI membership in 2014 were similarly unreasonable and economically infeasible. In January 2014, Reynolds preliminarily quoted MVSC fees of $270-$330 per month, per dealer rooftop, for the packages MVSC requested. (MVSC SOFR ¶¶ 27-28.) Reynolds later quoted fees of $340-$395 to add two additional packages with "writeback" capabilities—i.e., the ability to write or input data back into the DMS—that MVSC had requested. (Id. ¶ 29.) MVSC witnesses have acknowledged that MVSC requested more services than it strictly needed to process EVR transactions, but MVSC insists that it needed those services "to innovate [its] product offerings." (Id. ¶¶ 27, 30.)

MVSC did not initially move forward in January 2014, but reached out again in August to request quotes for three different software packages; Reynolds responded with rates ranging from $280 to $420 depending on the package. (See id. ¶¶ 31-33.) Negotiations were unsuccessful: MVSC informed Reynolds that state regulations—which capped the fees MVSC could charge—rendered Reynolds's proposed fees economically unfeasible. (Id. ¶ 35.) Reynolds responded that it charged "like" pricing for "like" integration—in other words, that its policy was to charge the same price for the same service across various categories of vendors. (Id.) MVSC disputes that, however, and points to evidence that Reynolds charged EVR providers significantly more than non-EVR vendors for the same access. (Id.) MVSC told Reynolds in November 2014 that it did not intend to move forward with RCI. (Id. ¶ 36.)

7. Other EVR providers

At least one of MVSC and CVR's competitors have evidently accepted Reynolds's terms: Title Technologies, LLC ("TitleTec") has participated in RCI since 2010.²⁹ (MSVC SOFR ¶ 37.) TitleTec competes with CVR in all six states where TitleTec operates: Florida, Georgia, Virginia, Maryland, Pennsylvania, and South Carolina. Reynolds perceives TitleTec as one of CVR's "two major competitors"—the other is DealerTrack RTS. (Id.) DealerTrack, which does not participate in RCI, provides EVR services in 15 states and competes with CVR in 12 states, including California, Illinois, Wisconsin, and Virginia. (Id. ¶¶ 23-24.)

29 Other, smaller competitors have also joined RCI, including AIB, Inc. (MVSC SOFR ¶ 38.)

8. Alleged evidence of conspiracy to deny access to certified integration

MVSC cites several pieces of evidence as suggestive of a conspiracy between Reynolds and CDK to deny MVSC access to certified integration. Most notably, MVSC COO Nemelka testified that in either December 2015 or December 2016, Christopher Morris—at that time, a CDK employee, but formerly a CVR Board Member on behalf of Reynolds—reached out to recruit MVSC for CDK's 3PA program. (See Nemelka Dep. [1069-5] at 375:14-376:13.) Nemelka was skeptical; he "expressed some frustration" to Morris about how MVSC had "been denied multiple times" in the past. (Id. at 376:15-20.) Morris responded that there were "potentially,

new opportunities or new doors that may be open that may allow [MVSC] the opportunity to participate in 3PA." (Id. at 376:21-24.)

According to Nemelka, Morris separately told him that in CVR board meetings, CDK, Reynolds and CVR had "talked about" MVSC and "discussed trying to keep [MVSC] from being competitive, specifically in California." (Id. at 377:5-18.) Nemelka continued: "I think he actually used the word 'closed category' at one time" because CVR "would allow no vendor in the [3PA] program who competed with CVR in any state that they cared about." (Id. at 377:21-378:2.) Though Nemelka did not remember the exact wording, he understood Morris to be saying that if CVR operated in a particular state, "they"—meaning "Reynolds, CDK and CVR"—"had a closed category whereby they would not authorize other EVR vendors to participate in the 3PA program." (Id. at 378:3-8, 379:12-17; see also DRPSOF ¶ 101.)

Emails among CDK, Reynolds, and CVR provide other evidence of collusion between Reynolds and CDK to block MVSC from accessing certified integration. For example, in January 2014, John Roeder of CVR emailed Jon Strawsburg of Reynolds (who sits on CVR's board) to offer himself as an EVR expert for customers and prospective customers visiting Reynolds's booth at the NADA convention. Roeder stated: "Together we can also share with clients/prospects the unique value a[] [Reynolds]/CVR solution brings to dealerships around data integrity and security. As you know[,] Reynolds & Reynolds receives a revenue share on each CVR transaction which represents millions of dollars annually. We're [] all chasing [r]evenue and including CVR can have significant value" to Reynolds. (DRPSOF ¶ 102.)

MVSC asserts that several internal documents following CVR's April 2015 acquisition of AVRS—a California-based EVR operator—suggest a conspiracy to "deny MVSC access to 'direct DMS integration.'" (Id. ¶ 104.) On October 16, 2015, Mark Kithcart, Vice President of Sales and Marketing at AVRS, sent proposed talking points to an internal group. "The focus and big difference" between CVR/AVRS and MVSC, according to Kithcart, was that MVSC "does not and WILL NOT have direct DMS integration." (See [1089-21].) According to Kithcart, "the future is not having a stand-alone DMV product. The future is having an integrated and secure DMV product that can respond and act intelligently based on the data available. The data that [MVSC] will NEVER have access to." (Id.) The next month, when Strawsburg from Reynolds reached out to Herbers—the CDK executive responsible for overseeing CVR at that time—for a "quick summary of the advantages of CVR/AVRS" as compared to MVSC, Herbers sent talking points drafted by Kithcart that contained the same language from Kithcart's October email. (See [1069-131] (email chain between Strawsburg and Scott repeating language that MVSC "WILL NOT have direct DMS integration" with Reynolds and would "NEVER have access" to certified data through Reynolds).)³⁰ According to Reynolds, Kithcart—who drafted the talking points for marketing AVRS—had no authority over RCI admissions. (DRPSOF ¶ 104.) 9. MVSC's use of independent integrators

30 A document called "CVR/AVRS Sales Training Plan" dated March 3, 2016 suggests that a "Reynolds Account Manager" talking to a "Dealer-principle [sic] or Controller" could tell the prospective client that MVSC "does not and WILL NOT have direct DMS integration with Reynolds & Reynolds" and that MVSC was "actually a hostile integration." (See [1069-132].) Reynolds responds that the "final version" of that marketing plan did not include the statement about DMS integration. (See DRPSOF ¶ 104.)

MVSC used only one independent integrator on the Reynolds DMS—Authenticom—and used it for at least four dealers. (MVSC SOFR ¶ 55.) By 2013, MVSC recognized that "Reynolds seems determined to lock out users they consider 'un-certified,'" and MVSC therefore planned to "start pursuing alternatives to Authenticom," and in fact stopped using Authenticom's services the following year. (Id. ¶ 56.) MVSC used a broader set of independent integrators on the CDK DMS and continued doing so until 2017. (See id. ¶ 58.) In 2016, CDK deployed various security measures to prevent unauthorized access to its DMS. (DRPSOF ¶ 110.) MVSC converted the majority of its CDK dealers to manual import methods prior to 2017. (MVSC SOFR ¶ 58.)

10. Messaging

MVSC cites evidence that in marketing CVR, Reynolds targeted MVSC's lack of certified access to data. For example, MVSC points to training materials for a Reynolds-CVR initiative to "re-conquest" (i.e., win back business from) big dealers in California, which included talking points highlighting that "Security is a MUST today—and only direct & certified integration can guarantee this." (DRPSOF ¶ 109.) The talking points also noted that AVRS (CVR's acquisition) was "[t]he only certified & secure Reynolds & Reynolds DMV vendor in California." (Id.) MVSC also cites a marketing document labelled "AVRS DMVDESK COMPARISON" that describes AVRS as "[d]irectly integrated with CDK/Reynolds & Reynolds" and "CDK/RCI Approved," comparing favorably with MVSC, which "Utilizes 3rd party" and is "Not CDK/RCI Approved." (Id.; see also MVSC SOFR ¶ 61 (describing Reynolds marketing efforts).) Reynolds disputes any suggestion that CDK, CVR, and Reynolds jointly marketed MVSC's lack of certified integration or that Reynolds marketed MVSC as "insecure," but cites no evidence to support its claim. (DRPSOF ¶ 109.) Indeed, Reynolds responded to the evidence discussed above—where CVR officials claimed MVSC would never have access to certified integration—by characterizing the communications are mere "proposed talking points that Reynolds could use to market AVRS's services," a point somewhat in tension with its position that the companies never jointly marketed MVSC's lack of certified integration. (Id. ¶ 104.)

11. California

As reflected in the following chart, from 2014 until 2019, MVSC's market share in California steadily grew, while CVR's share held (mostly) steady (see MVSC SOFR ¶ 62):

Year	2014	2015	2016	2017	2018	2019
MVSC	36.8%	48.6%	53.8%	58.2%	57.1%	61.1%
CVR	34.2%	38.6%	35.0%	30.5%	31.4%	29.8%

MVSC announced in April 2014 that it was the number one EVR provider in California among dealers and noted in various 2015 emails that MVSC was significantly out outperforming CVR in the California market. (Id. ¶ 63.)

Despite those metrics, MVSC claims that its lack of access to data through "certified" integration or independent integrators

limited its growth in California or caused it to lose customers. In a declaration cited by MVSC, Nemelka reports that in "conversations with dealer customers and MVSC sales staff," he learned that several dealer groups had cited MVSC's lack of certified or automated access to data as primary or significant reasons for leaving or not doing business with MVSC. (See DRPSOF ¶ 112.) MVSC also cites Dr. Israel's opinions that (1) MVSC has been able to succeed in California despite Defendants' anticompetitive conduct because it offers a superior product; and (2) MVSC's market share in California would have been higher but for the anticompetitive conduct. (Id.) Reynolds emphasizes, by contrast, that "MVSC has been the dominant EVR provider in California from 2014 to the present, without exception, commanding more than 60% market share in California as of June 2019." (Id.)

12. Virginia

In 2014, MVSC obtained a contract with the State of Virginia to begin doing business in that state but did not begin processing transactions until June 2017. (DRPSOF ¶ 115.) The parties dispute whether MVSC's lack of integrated data access delayed its entry into Virginia. MVSC cites Nemelka's declaration that in Virginia, MVSC's lack of "certified" access to data delayed its ability to roll out a viable EVR solution in the market until it developed Electron (its workaround for near real-time data access) in late 2017. (Id.) Reynolds responds that the delay was caused by MVSC at first and later by the state, which took several months to complete its testing. (Id. ¶ 116.) Reynolds cites emails showing MVSC's development team did not begin working on the user interface for Virginia until June 2016 and was working to complete its own testing in August 2016, and an email showing that MVSC was at the Virginia DMV for testing in March 2017. (Id.) It was that slow pace, not lack of data access, which delayed MVSC's progress in Virginia, Reynolds contends.

Reynolds also argues that MVSC did not need certified integration to develop its Virginia EVR solution, citing an MVSC email identifying manual reporting as an option for data access. (Id.) The email suggested that manual reporting could be advertised as a "security feature" that would give dealers greater control over their data. (Id.) That email confirms, Reynolds argues, that Reynolds's free manual reporting was sufficient for MVSC to process EVR transactions before it developed Electron. (Id.)

13. Wisconsin

MVSC has attempted to expand into Wisconsin since 2013. (MVSC SOFR ¶ 66.) Again, the parties dispute whether MVSC's lack of access to data through "certified" integration or independent integrators has impeded MVSC's entry into Wisconsin. (DRPSOF ¶ 118.)

MVSC submitted a formal application for approval to the Wisconsin EVR program in May 2014. (MVSC SOFR ¶ 67; [957-45].) The Wisconsin DMV denied the application in August, citing the "extensive resources" needed from the DMV for "start-up programming and testing" and the lack of benefit to "either DMV or our customer, the Wisconsin motorist." (MVSC SOFR ¶ 67.) MVSC disputes that those were the only reasons for the denial. (Id.) Wisconsin officials seemed initially enthusiastic about MVSC entering the Wisconsin market, MVSC contends, citing, for example, an internal email from an MVSC regional manager who summarized that MVSC personnel had a "very welcoming meeting with the Wisconsin DMV" who appeared "very interested" in MVSC's offerings. (Id. ¶ 66.) MVSC also stresses that the DMV's concern about technical support "would likely encompass aspects of MVSC's service offering that require data import from dealers' DMSs." (Id. ¶ 67.)

Still, MVSC's CEO conceded that he was not aware that any Wisconsin official expressed concern regarding MVSC's level of integration. (Id. ¶ 70.)

In February 2017, staff at Wisconsin's DMV informed MVSC that while the DMV was not accepting new vendors into its EVR program when MVSC initially applied, it hoped to implement new vendors in 2019. (Id. ¶ 69.) As of the parties' briefing, Wisconsin had not approved a new EVR provider since 2005. (Id. ¶ 68.) As of 2015, six other EVR providers had expressed interest in entering the Wisconsin EVR market but were rejected. (Id.)

14. Illinois

Four companies, including CVR and MVSC, provide registration services in Illinois. (MVSC SOFR ¶ 78.) MVSC applied to participate in Illinois's EVR program in 2013 and was subject to a two-step approval process. (Id. ¶¶ 72-73.) First, MVSC needed to demonstrate its ability to process vehicle registration renewals (a separate task from processing registrations) meaning that MVSC was required to develop and get state approval for a renewals app—a process that MVSC did not complete until 2015. (Id. ¶ 73.) Next, Nemelka testified, after MVSC had processed "enough" renewals for the State to "get comfortable" with MVSC, MVSC would then develop its EVR (i.e., registrations) app for the State to test and approve. (Id.)

According to MVSC, it obtained "formal authorization" to develop its EVR app by early 2015. (DRPSOF ¶ 113.) But Reynolds points out that MVSC's 2015 Illinois contract states that MVSC was an approved provider "for purposes of renewing vehicle registrations," not for initial EVR transactions. (Id.) Reynolds also cites an email Nemelka wrote to an Illinois official a day after Illinois signed the 2015 contract in which Nemelka said that MVSC was "probably 6 months away" from being approved for EVR processing. (Id.) Once MVSC developed its Illinois EVR app, it had to wait for the State to test and approve the app, a process that was delayed due to the State's budget and "resource constrain[ts]." (MVSC SOFR ¶ 75.) Illinois did not begin testing MVSC's EVR app until October 2017; but by the end of December 2017, MVSC had been approved. (Id. ¶ 76.)

As in other markets, MVSC asserts—but Reynolds disputes—that MVSC's lack of data integration delayed its entry into Illinois. (See DRPSOF ¶ 114.) Reynolds attributes delays in Illinois to the State's lengthy two-step approval process and points out that Illinois never expressed concerns about MVSC's DMS integration. (Id.) Reynolds also claims that the absence of independent integrators had no effect on MVSC's Illinois efforts because those integrators could not provide the "near real-time" access that MVSC claims it needs in Illinois. (Id.)

B. Sherman Act Section 1 Conspiracy

MVSC alleges that Reynolds and CDK conspired to (1) block MVSC's access to certified data integration through RCI and 3PA; and (2) block MVSC's access to independent data integration. The second part of the alleged conspiracy survives summary judgment for the same reasons discussed above. See Section II.B, supra. MVSC maintains that there is likewise ample evidence that MVSC suffered injury and damages from Reynolds's agreement with CDK to "close" its systems to CVR's competitors, as discussed further below.

1. Element 1: Agreement

Reynolds argues that summary judgment is warranted on MVSC's certified integration claim because MVSC cannot marshal sufficient evidence: (1) to show that Reynolds conspired with CDK to deny MVSC access to data it needed to compete, or (2) to show that Reynolds (rather than

CDK) refused to deal with MVSC. After resolving material factual disputes in MVSC's favor and viewing the record holistically, the court holds that MVSC has presented sufficient evidence that Reynolds and CDK had a "conscious commitment to a common scheme designed to achieve an unlawful objective." Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 768, 104 S.Ct. 1464, 79 L.Ed.2d 775 (1984). Reynolds's challenge to MVSC's evidence on this score asks MVSC to produce a "single item of evidence" that "points equivocally to conspiracy" and would require the court to "weigh conflicting evidence (the job of the jury)"—practices the courts have called a common "trap" in conspiracy cases. See High Fructose Corn Syrup, 295 F.3d at 655.

a. Conspiracy

Reynolds argues that MVSC cites no evidence that Reynolds unlawfully conspired with CDK to block MVSC's access to data, much less evidence that tends to exclude the possibility of independent action. But Reynolds largely ignores the strongest evidence in MVSC's favor: draft marketing materials stating that MVSC "does not and WILL NOT have direct DMS integration" with Reynolds, and Nemelka's conversation with a former CVR board member who suggested Reynolds and CDK viewed EVR as a "closed category." (See Nemelka Dep. at 377:21-378:8.) These communications go well beyond "the unremarkable fact that Reynolds and CDK occasionally met and/or communicated about CVR, their joint venture." (Reynolds's MVSC MSJ at 15.) Viewed in the light most favorable to MVSC, they support the existence of a conspiracy rather than independent action.

Reynolds insists that the marketing emails are not evidence of a conspiracy because no one on the email chain had any "authority relating to MVSC's potential admission into RCI; they were concerned only with EVR marketing." (Reynolds's MVSC Reply [1121] at 10.) But those individuals' (lack of) authority over Reynolds's alleged policy to exclude MVSC from RCI says nothing about their knowledge of it. Reynolds also notes that the statement—that MVSC "does not and WILL NOT have direct DMS integration"—is factually true, because MVSC did not join RCI. (Id. at 9-10; see also Section III.A.6, supra.) But Reynolds construes the statement in the light most favorable to itself. Viewed instead in MVSC's favor, the statement that MVSC "WILL NOT" have access to data via RCI suggests a forward-looking agreement not to allow MVSC into the program. Reynolds also notes that the proposed language was not used in its final sales training plan. But the court has no information about why the language was removed. It will be the jury's job to interpret and weigh the emails.

In addition to the marketing emails, MVSC relies on Nemelka's testimony about a conversation he had with Christopher Morris—at that time, a CDK employee, but formerly a Reynolds employee and CVR Board member on behalf of Reynolds.³¹ According to Nemelka, Morris told

31 In a footnote in its reply brief, Reynolds asserts that Nemelka's testimony is inadmissible hearsay. (Reynolds's MVSC Reply at 8, n.7.) Reynolds waives this argument by failing to develop it. Reynolds has not demonstrated that the testimony would be inadmissible, nor does it discuss potentially applicable hearsay exceptions, such as co-conspirator statements under Rule 801(d)(2)(E). Reynolds also claims that Nemelka's testimony about what Morris told him is "self-serving deposition testimony ... insufficient to withstand summary judgment." (Id.) But as the Seventh Circuit has repeatedly emphasized: "the term 'selfserving' must not be used to denigrate perfectly admissible evidence through which a party tries to present its side of the story at summary judgment." Hill v. Tangherlini, 724 F.3d 965, 967 (7th Cir. 2013). The court will consider testimony that is "based on [Nemelka's] personal knowledge" and that sets forth "specific facts." See id. at 968.

him that CDK, Reynolds and CVR "talked about" MVSC during CVR board meetings and "discussed trying to keep [MVSC] from being competitive, specifically in California." (Nemelka Dep. at 377:5-18.) While the Morris conversation may not be direct evidence of an agreement to deny MVSC access to RCI—recall that Morris called Nemelka to discuss MVSC's joining CDK's 3PA program—the conversation provides circumstantial evidence that Reynolds colluded with CDK to "keep [MVSC] from being competitive." (Id.) Indeed, MVSC argues that Reynolds knew the rates it proposed to MVSC were unworkable. When MVSC sought to join RCI in 2014, Nemelka explained to Reynolds—with citations to California and Illinois regulations—that state caps on EVR rates made Reynolds's proposal economically unfeasible. (MVSC SOFR ¶ 35.)

Reynolds also concedes that CDK—which admitted internally that it refused to deal with MVSC because MVSC competed with CVR—acted in a manner consistent with such an agreement. CDK literally or constructively refused to deal with MVSC five times during 2015 and 2016. (See Section III.A.5, supra (detailing CDK's outright refusals and economically unworkable deal terms).) It is not a far stretch to infer that Reynolds's proposed fees were also an attempt to protect CVR. A reasonable jury could accordingly infer that Reynolds and CDK agreed not to allow MVSC into their certified access programs because of the competitive threat posed by MVSC.

Reynolds cites evidence to the contrary but has not established the absence of disputed facts. First, Reynolds argues that MVSC cannot demonstrate a conspiracy because, when MVSC attempted to join RCI and 3PA, Reynolds's and CDK's responses were "not parallel or coordinated in the least." (Reynolds's MVSC MSJ at 13.) "[W]hile CDK refused to allow CVR competitors into 3PA," Reynolds argues, CDK "was well aware that Reynolds had an opposite, open policy of admitting CVR competitors into RCI." (Id.) Reynolds's evidence on this score is thin. It cites a single 2014 CDK PowerPoint titled "Third Party Access Strategy"—"Strategy 'Refresh.'" (See [958-42] at CDK-0002272.) One slide in the deck contains the line, "Reynolds and Reynolds Certified program ... No Categories/Category Restrictions (ex: CVR competitors included)." (Id. at CDK-0002380.) Whether Reynolds was, in fact, open to MVSC is a jury question for the reasons explained above.

Reynolds also asserts that "MVSC itself acknowledged this critical difference" between Reynolds's posture and that of CDK, quoting a September 2016 email from Nemelka to a dealer contact, where Nemelka says that "[u]nlike Reynolds," CDK "effectively won't even let us participate in the [3PA] program.'" (See Reynolds's MVSC MSJ at 14) (quoting [958-43]).) But read in context, the email is not so supportive. True, the most aggressive criticism in Nemelka's email is aimed at CDK's anticompetitive behavior. But Nemelka is writing his contact to ask if the contact's employer—a "Reynolds dealership group"—would consider sending a letter to the California DMV complaining about Reynolds's behavior. (Id.) To that end, Nemelka writes: "As you know, Reynolds charges an unreasonable amount to participate in their RCI program and they keep chasing guys like us around trying to keep us from getting the data out of the DMS if we aren't a RCI participant." (Id.) Second, Reynolds argues that CDK had a strategy to deal with certified integration for EVR providers that was "completely different" from that of Reynolds. Unlike Reynolds, CDK expressly rejected MVSC's first two requests to join the 3PA program and later demanded a percentage-of-revenue fee of 35 percent, as compared to what Reynolds calls its own "fixed, nondiscriminatory dollar amounts." (Reynolds's MVSC MSJ at 1, 14.) But "[s]imultaneous action is not a necessary element of parallel conduct" and MVSC is not required to show that Defendants "restricted supply in an identical manner." In re Turkey Antitrust Litig., No. 19 C 8318, 642 F.Supp.3d 711, 722-28 (N.D. Ill. Nov. 21, 2022). In any event, Reynolds and CDK's challenged actions are similar: according to MVSC, both proposed fee structures MVSC claims were cost prohibitive. Cf. In re Broiler Chicken, 290 F. Supp. 3d at 792 ("It is more than plausible that conspirators would leave the precise means of cutting production up to each conspirator, where multiple options would accomplish the intended goal.").

Reynolds is on firmer ground with its third contention: that MVSC's theory "makes no sense," given that Reynolds declined for years to certify its own joint venture, CVR, for participation in the RCI program because doing so would have required RCI to give CVR preferential treatment. (Reynolds's MVSC MSJ at 20.) As Reynolds frames it: "[n]o rational actor would enter into a conspiracy to assist its minority-owned joint venture and then not actually take action to assist it." (Id.) The argument has some traction, but although Reynolds did not complete an RCI interface for CVR until August 2017 (see id.), there is evidence that Reynolds gave CVR a product superior to the one it made available to its competitors, including additional interfaces and writeback capability. (See MVSC MSJ Opp. [1073] at 24 n.10; [1090-46] (Bob Brockman: "Obviously we want to help CVR sell stuff. Therefore I am approving this request.") Further, the record reveals that Reynolds listed CVR as participating in its certified integration program (RCI) as of January 2014 even though CVR did not have real-time integration at that point. (See [1090-42] at 4; see also [1090-79] at 2 (June 2015 email from Honda dealer telling MVSC they are staying with CVR "on the grounds that they are Reynolds & Reynolds 'certified', as [Reynolds] owns a part of AVRS.").)

Fourth, Reynolds maintains that the conspiracy "makes no economic sense" because CVR's market share "steadily declined year over year while MVSC's went up to over 60% in California." (Reynolds's MVSC MSJ at 8.) While true on its face, there is a material factual dispute about how to understand those facts. Klein, MVSC's expert, analyzed MVSC's growth in market share over the relevant period. He observed that MVSC's growth began to decline sharply in 2015, just when Defendants' anticompetitive conduct intensified—most notably with the signing of the 2015 Agreements. (MVSC SOFR ¶ 62.) In other words, though MVSC's market share continued to grow, its rate of growth slowed. The cases Reynolds cites, on the other hand, involved plaintiffs whose market share grew "constantly" or "steadily" during the alleged conspiracy period. See, e.g., King v. Idaho Funeral Serv. Ass'n, 862 F.2d 744, 745-46 (9th Cir. 1988) (affirming summary judgment for defendants). More fundamentally, the Seventh Circuit has warned courts to "distinguish between the existence of a conspiracy and its efficacy." High Fructose Corn Syrup, 295 F.3d at 656. Put differently, Reynolds may be liable for violating the Sherman Act even if its conspiratorial conduct was unsuccessful. Fifth, Reynolds notes that at least five other EVR providers have joined RCI, one of whom—TitleTec—Reynolds sees as major competitors.³² (See Reynolds's MVSC MSJ at 22-23.) Reynolds maintains that, if MVSC believes that the "conspiracy only targeted MVSC because it posed the strongest competitive threat to CVR," TitleTec and DealerTrack's strong market positions belie that notion. (Id. at 22.) Reynolds, however, ignores evidence suggesting that MVSC was a particularly significant threat because it competed with CVR in California and Illinois, the two states where CVR generated the most revenue. (MVSC MSJ Opp. at 18.) And TitleTec's 30(b)(6) witness testified that the rates Reynolds proposed to MVSC would have kept TitleTec out of the California and Illinois markets, too. (Id. at 22.)

32 MVSC's other major competitor, DealerTrack, was invited to join, but declined. (See MVSC SOFR ¶ 23.)

Sixth, Reynolds argues that it had no "rational incentive to participate in a lopsided agreement and take on the risk of an unlawful scheme" given that CDK's stake in MVR is four times as large. (Reynolds's MVSC MSJ at 23.) But this case is unlike Washington County Health Care Authority, Inc. v. Baxter International, Inc., where the court rejected a conspiracy claim based on a bogus recall of saline because "[i]t makes little sense" that one conspirator "would agree to recall eight times as much product," "incurring substantially greater costs despite roughly equal market share." 328 F. Supp. 3d 824, 832, 837 (N.D. Ill. 2018). In this case, unlike in Baxter, Reynolds incurred no "high upfront costs," nor did it invite regulatory "scrutiny" by participating in the alleged conspiracy. Id. at 832. Indeed, so long as Reynolds stood to make more from its stake in MVR than it would lose in RCI revenue from blocking MVSC, the conspiracy did make economic sense.

Seventh, Reynolds argues that any evidence that it disparaged MVSC while marketing CVR or communicating with state regulators is not actionable under the antitrust laws because those communications are protected by the First Amendment. (See Reynolds's MVSC MSJ at 17-18) (citing cases).) In its reply brief, Reynolds further claims that because MVSC's expert calculated damages for all alleged misconduct—including so-called marketing interference in addition to "blocking" efforts—and admitted he cannot disaggregate damages, his opinions must be excluded and MVSC is left with no proof of damages. (Reynolds's MVSC Reply at 19.)

Reynolds is correct that disparaging or even "[f]alse statements about a competitor—without more—do not give rise to an antitrust violation." Hytera Commc'ns Corp. Ltd. V. Motorola Sols., Inc., 623 F. Supp. 3d 857, 879-80 (N.D. Ill. 2022). MVSC emphasizes, however, that it is not pursuing a theory that Defendants' marketing or lobbying efforts were themselves illegal. Instead, MVSC cites "[e]vidence that CVR relied on its exclusive access to data" in marketing and lobbying to support the inference that "such access was critical and that Defendants' denial of that access affected MVSC's ability to obtain regulatory approval." (MVSC MSJ Opp. at 31.) When a defendant claims its speech immunizes it from liability, the court must "first identify any conduct that is immunized" and then "consider the evidence of the remaining challenged conduct in the aggregate to see if it is sufficient to support antitrust liability." Mercatus Grp., LLC v. Lake Forest Hosp., 641 F.3d 834, 839 (7th Cir. 2011). Notably, "boycotts and agreements not to distribute certain products are the types of enforcement mechanisms

that may render speech actionable under the antitrust laws." Id. at 851 (citation omitted).

The court has determined that sufficient evidence supports MVSC's group boycott theory based on Reynolds's exclusion of MVSC from the RCI program and its effort to block independent integrators. Thus, even ignoring arguably immunized conduct, MVSC still presents facts "sufficient to support antitrust liability." Id. at 839; cf. Hytera Commc'ns, 623 F. Supp. 3d at 880 (false marketing statements about competitor were "part of a larger patten of behavior, intended to maintain [the defendant's] monopoly"); Nexstar Broad., Inc. v. Granite Broad. Corp., No. 1:11-CV-249, 2012 WL 2838547, at *8 (N.D. Ind. July 9, 2012) (denigrating speech could not state antitrust claim on its own, but would be part of a "course of conduct" that showed the defendant's "intent to monopolize the television advertising market and destroy competition"). Finally, the court rejects Reynolds's disaggregation-of-damages argument as waived.³³

33 Reynolds's argument concerning Klein's inability to disaggregate damages attributable to "market interference" was raised for the first time in reply and inadequately developed. See, e.g., O'Neal v. Reilly, 961 F.3d 973, 974 (7th Cir. 2020) ("[W]e have repeatedly recognized that district courts are entitled to treat an argument raised for the first time in a reply brief as waived.").

b. Refusal to deal

According to Reynolds, it never refused to deal with MVSC, but instead proposed "nondiscriminatory fees" comparable to "what Reynolds offered other EVR providers in RCI." (Reynolds's MVSC MSJ at 1.) MVSC walked away from those discussions in 2014. According to Reynolds, its policy was to quote "like" pricing for "like" integration to EVR providers, which is what it did with MVSC. MVSC makes two main points in response.

First, MVSC emphasizes that CDK explicitly refused to allow MVSC access to its 3PA program, which Reynolds had conceded. According to MVSC, CDK's refusal to deal with MVSC "is attributable to Reynolds as a co-conspirator," so "[e]ven if Reynolds would have been willing to defect from the conspiracy and give MVSC RCI access at a high enough price, that would not absolve Reynolds of liability." (MVSC MSJ Opp. at 21.) That argument is a bit perplexing. True, the "general right to refuse to deal with competitors" applies only to unilateral refusals. See Buccaneer Energy (USA) Inc. v. Gunnison Energy Corp., 846 F.3d 1297, 1309 (10th Cir. 2017). And true, "an antitrust defendant is jointly and severally liable for the acts of its co-conspirators." In re Uranium Antitrust Litig., 552 F. Supp. 518, 522 (N.D. Ill. 1982). But MVSC's argument only applies if Reynolds and CDK are co-conspirators. And as a practical matter, if MVSC were to concede that Reynolds's proposed RCI prices were reasonable—in other words, did not constitute an effective refusal to deal—that would seem to undermine the very foundation of the alleged conspiracy. MVSC suggests, but does not really explain, that Reynolds might be viewed as simply "defecting" from the conspiracy by "giv[ing] MVSC RCI access at a high enough price." (MVSC MSJ Opp. at 21.) The court is not persuaded.

The court does, however, credit MVSC's second argument: that there are material factual disputes concerning whether Reynolds effectively refused to deal with MVSC, too. Construing the record in the light most favorable to MVSC, a reasonable jury could find that Reynolds's proposals constituted constructive refusals to allow MVSC to participate in RCI. Of course, Reynolds was under no obligation to offer MVSC especially generous prices

or to "deal under terms and conditions that [MVSC] find[s] commercially advantageous." Pac. Bell Tel. Co. v. linkLine Commc'ns, Inc., 555 U.S. 438, 450, 129 S.Ct. 1109, 172 L.Ed.2d 836 (2009). But in some instances, a party's proposed terms may be so unreasonable that the offer amounts to an effective refusal. See, e.g., Re/Max Int'l, Inc. v. Realty One, Inc., 173 F.3d 995, 1014-15 (6th Cir. 1999) (imposing "additional costs" on "boycotted entit[ies]" that were "economically unsustainable" constituted refusal to deal); Del. & Hudson Ry. Co. v. Consol. Rail Corp., 902 F.2d 174, 179-80 (2d Cir. 1990) ("[T]here need not be an outright refusal to deal.... It is sufficient if the terms of the offer to deal are unreasonable.").

Reynolds argues that the rates it offered MVSC were the same rates that two other EVR providers, TitleTec and AIB, agreed to pay Reynolds. (Reynolds's MVSC MSJ at 11-12.) But MVSC disputes that those companies are proper comparators. It emphasizes that the prices Reynolds offered to MVSC were far higher than the prices quoted to other, non-EVR, application providers for the same interfaces. (See MVSC MSJ Opp. at 22.) More importantly, TitleTec and AIB do not compete in California (MVSC's primary market) or Illinois—a large market where MVSC was actively planning to expand. (See Section III.A.7, supra.) A reasonable jury could conclude that because regulations—most notably price caps—vary by state, the market must be analyzed on a state-by-state basis. TitleTec's and AIB's decision to pay Reynolds's proposed rates therefore does not establish that the rates were economically feasible for EVR operators like MVSC doing business in states where the prices EVR operators could charge were subject to regulatory caps. Indeed, TitleTec's 30(b)(6) deponent testified that the fees Reynolds charged for access to RCI would have prevented TitleTec from competing in California and Illinois. (See id. at 30.)

c. Independent integration

Reynolds first argues that there is no evidence it conspired with CDK to block MVSC's access to data through independent integrators, an argument which the court has already rejected. See Section II.B, supra.

Reynolds separately claims that MVSC did not actually rely on independent integration services because—except in its dealings with four dealers in 2013—an MVSC employee testified that it never used Authenticom or any other independent integrator to access data from a Reynolds's DMS. (MVSC SOFR ¶ 55.) MVSC has presented evidence, however, that it stopped using Authenticom for integration after 2013 precisely because of Reynolds's aggressive blocking efforts, which forced MVSC to develop a manual workaround. (See MVSC MSJ Opp. at 15-16; [958-47] ("[S]ince Reynolds seems determined to lock out users they consider 'un-certified,' we will start pursuing alternatives to Authenticom."); [957-40] (MVSC "work[ed] on a parallel solution" for manually importing data "when Reynolds started clamping down and a few of our clients were on Authenticom").

2. Element 2: Unreasonable restraint

The parties dispute whether MVSC's § 1 claim is subject to a per se analysis or the rule of reason. The rule of reason is the presumptive "standard framework" under § 1. Agnew v. Nat'l Collegiate Athletic Ass'n, 683 F.3d 328, 335 (7th Cir. 2012). The court already determined it applies to the AutoLoop/Dealer case, above, see Section II.B.2, supra, and concludes that a rule-of-reason analysis applies here, too. See, e.g., Ind. Fed'n of Dentists, 476 U.S. at 458-59, 106 S.Ct. 2009 (dental association policy requiring members to withhold x-rays from dental insurers was subject to the rule-of-reason analysis, notwithstanding

that policy resembled a group boycott, which is unlawful per se).

3. Element 3: Injury

Reynolds next asserts that MVSC cannot establish that the Reynolds-CDK conspiracy was a substantial factor in injuring MVSC. "[P]laintiff must establish, with a fair degree of certainty, that the violation was a material element of, and substantial factor in producing, the injury." Greater Rockford Energy and Tech. Corp. v. Shell Oil Co., 998 F.2d 391, 401 (7th Cir. 1993) (affirming summary judgment where intervening factors barred plaintiff from establishing that alleged antitrust violations were the but-for cause of plaintiff's injuries). Reynolds argues that, to the extent MVSC suffered harm, it was due to other causes. As explained here, the court agrees with Reynolds, but only with regard to one of the states in which MVSC claims harm.

a. California

Reynolds maintains that MVSC suffered no harm in California because, before California became a point-of-sale state in 2019, MVSC could obtain the data it needed through Reynolds's free manual reporting tool, and after that time, MVSC used its Electron program to augment the manual reporting tool. Reynolds also points to several undisputed facts: (1) MVSC used manual reporting for years to access data in most DMSs, not just Reynolds's and CDK's; (2) MVSC told customers that manual reporting was sufficient; and (3) Reynolds never blocked MVSC from using its manual reporting, nor did it block MVSC from using the Electron tool. Reynolds emphasizes that MVSC—using Reynolds's manual reporting tools—overtook CVR as the leading EVR provider in California as early as April 2014. As Reynolds puts it, MVSC growth "has continued without interruption, all without certified integration or the use of hostile integrators to access the Reynolds DMS." As noted earlier, MVSC's California market share rose from less than 37 percent at the start of the alleged conspiracy in 2014 to more than 61 percent by 2020. (Reynolds's MVSC MSJ at 25-26.)

MVSC nonetheless insists that its lack of access to automated integration was a material factor in the loss of California customers. MVSC first cites Nemelka's declaration, which presents a list of dealerships that listed MVSC's lack of "certified" or automated access to data as a primary or significant reason to stop doing business with MVSC. (See [1090-61] ¶¶ 12-16.) MVSC also reviews a variety of specific examples, including an email from a former customer telling MVSC that though it had "been a class act," the dealer was switching to CVR because "the integration piece is what is dictating the relationship." (MVSC MSJ Opp. at 26.) Another dealership group left MVSC for AVRS because management wanted it to use only "vendors that are part of the 3PA program" and "they want to take advantage of the 'features and integration' that come with" certified integration. (Id.; see, e.g., [1090-6] (MVSC employee forwarding October 2016 cancellation notice from dealer and reporting that dealer said it was switching to AVRS because AVRS is "on CDK system"); [1090-9] (similar letter from dealership in November 2016); see also Reynolds's MVSC Reply at 14 n.11 (listing others).)

Reynolds argues that "[a]ll of the MVSC emails purportedly conveying non-party dealer statements" are "inadmissible hearsay." (Reynolds's MVSC Reply at 14 n.11.) But this undeveloped argument, raised in a footnote of a reply brief, does not demonstrate that MVSC will be unable to present dealer statements about certified integration "in a form that would be admissible in evidence," FED. R. CIV. P. 56(c)(2), such as by calling as a witness the parties involved in the emails. See Mitsui

Sumitomo Ins. Co. v. Moore Transp., Inc., 500 F. Supp. 2d 942, 950-51 (N.D. Ill. 2007) (perfunctory, undeveloped, and skeletal arguments are waived).

In any event, at least some statements come directly from dealers. (See, e.g., [1090-10] (December 2016 email from Hyundai business manager telling MVSC: "We are leaving DMV desk to go to AVRS due to the import situation currently. I won't have to run a report to import all of our deals with AVRS ..."); [1090-79] (June 2015 email from Honda dealer telling MVSC they are staying with CVR "on the grounds that they are Reynolds & Reynolds 'certified,' as [Reynolds] owns a part of AVRS").) MVSC also provides Klein's opinion, who observed that MVSC's growth began to decline sharply in 2015, at a time when Defendants' anticompetitive conduct intensified. See Section III.B.1.a, supra (discussing Klein's findings).

Although not overwhelming, viewed in the light most favorable to MVSC, the evidence sufficiently establishes that Reynolds's alleged conspiratorial conduct materially and substantially produced MVSC's injury in the California market. See Rockford Energy, 998 F.2d at 401.

b. Virginia

As in California, Reynolds argues that MVSC was not harmed by the alleged conspiracy in Virginia because it was able to receive data through Reynolds's free manual reporting tools. However, MVSC presents evidence that it could not effectively use manual reporting until it developed Electron in late 2017. (DRPSOF ¶ 116.) MVSC was approved to enter the Virginia market by 2014 (id. ¶ 115), leaving several years of delay that a jury could reasonably attribute to Reynolds's allegedly conspiratorial conduct. Reynolds points out that MVSC processed transactions in Virginia a few months before Electron went live, but MVSC responds that those early transactions were extremely minimal. (See MVSC MSJ Opp. at 29.) MVSC's claim of injury in Virginia survives this motion.

c. Wisconsin

Reynolds argues—and the court agrees—that if a state prevented MVSC's entry into its EVR market, the state, not Reynolds, is the cause of MVSC's lost profits. See, e.g., In re Wellbutrin XL Antitrust Litig. Indirect Purchaser Class, 868 F.3d 132, 165 (3d Cir. 2017) (affirming summary judgment on causation grounds, explaining that "a regulatory or legislative bar can break the chain of causation in an antitrust case"); RSA Media, Inc. v. AK Media Grp., Inc., 260 F.3d 10, 15 (1st Cir. 2001) (affirming summary judgment on causation grounds where plaintiff was not excluded from market by defendant's threats, but by state regulatory scheme).

MVSC applied to the Wisconsin EVR program in May 2014. The Wisconsin DMV denied the application in August 2014 for two reasons:

Technical support needed from the DMV ... Adding any new title/registration vendor requires extensive resources from DMV for start-up programming and testing.

...

Lack of program expansion to benefit DMV and the Wisconsin consumer ... It is unlikely that adding a new vendor would increase the number of participants in the program, because Wisconsin dealerships have been processing titled applications electronically since legally mandated in 2007.... Adding an additional title/registration vendor without expanding the locations participating in the [automated registration] program offers no benefit to either DMV or our customer, the Wisconsin motorist. In fact, dealers and agents that switch vendors

require significant time from DMV to facilitate that process.

(See MVSC SOFR ¶ 67 (quoting letter to MVSC from Wisconsin DMV [957-45]).)

MVSC admits the content of the letter but insists that its lack of automated access to data on Reynolds's DMS hampered its efforts to satisfy regulatory requirements in Wisconsin. (MVSC MSJ Opp. at 30.) The only arguably relevant evidence MVSC cites in support, however, is a set of internal CVR notes summarizing a meeting with the Wisconsin DMV, in which CVR apparently told the DMV that "DMS interfaces are a huge part of dealer adoption of [EVR software] as they add to ease of use and less keystrokes.... DMV will not enjoy the ramifications of keystroke error due to double entry, when that bad data is being sent to DMV." (DRPSOF ¶ 118 (citing [1089-183] at 3).) This is just one isolated line in a long email summary of the meeting which took place in August 2016, long after the Wisconsin DMV rejected MVSC's application.

MVSC's other evidence, largely generic, does not overcome the specific reasons given by Wisconsin's DMV. For example, MVSC relies on testimony from Nemelka that he "observed that our lack of 'certified' access to data has hindered our efforts to obtain state approval" and a statement by CDK that integration is "key to securing DMV approval for many states." (Id. ¶ 118). None of that evidence speaks to how much weight, if any, Wisconsin's DMV ascribed to "certified" access.

Finally, MVSC relies on Klein's reply report, where he opines that the 2014 letter from the DMV does not "diminish[] the reasonable inference that concerns about DMS integration played a role in the state's approval process," because the state's concern about how much "technical support" MVSC would require might have been a subtle reference to data integration. (See Klein Reply Rep. [957-5] ¶ 42.) The DMV's letter says that "any" new EVR vendor would require "extensive resources from DMV for start-up programming and testing." (MVSC SOFR ¶ 67.) It does not allude to any challenges related to data integration, and to read such a concern into this letter is largely speculative.

In short, MVSC identifies no basis on which a reasonable jury could conclude "with a fair degree of certainty" that the Wisconsin DMV rejected MVSC's 2014 application because of unspoken concerns—fostered by Defendants—that MVSC could not provide an effective EVR solution. Rockford Energy, 998 F.2d at 401. To the contrary, the Wisconsin DMV provided specific reasons for its decision; none of them related to data integration. Consistent with its given reasons, the DMV rejected all six other EVR application during the same period and has not approved any new entrants since 2005. (See MVSC SOFR ¶ 68.) As the DMV told MVSC in 2017, it was "not accepting new vendors" into its EVR program when MVSC applied in 2014. (Id. ¶ 69.) MVSC's expert Mr. Klein conceded that if Wisconsin were in fact "closed to all new [EVR] vendors," then it is "pretty clear" that there would be no harm to MVSC from the alleged conspiracy. (Klein Dep. [957-6] at 130:7-11.) As the DMV itself told MVSC, it was closed to all new vendors when MVSC applied in 2014. Reynolds is entitled to summary judgment on MVSC's Sherman Act § 1 claim as it relates to the Wisconsin market.

d. Illinois

MVSC's argues that, absent the conspiracy, it would have entered the Illinois market in January 2015 rather than early 2018. Reynolds argues that it was Illinois's lengthy two-step evaluation process—not a conspiracy—that caused MVSC's delayed entry. First, MVSC had to develop and receive approval for an EVR renewal app

and then had to process "enough" renewals for Illinois regulators to "get comfortable with" MVSC before developing its registration app for testing and approval. (MVSC SOFR ¶ 73.) MVSC was approved for renewals in early 2015 (see [1090-19] at 16), but did not test an EVR app until October 2017 and was not approved until December. (MVSC SOFR ¶ 76.) The parties dispute at what point MVSC was formally authorized to develop its registration app. (See Section III.A.14, supra.)

Though Reynolds attributes all delays to Illinois officials, MVSC cites Klein's report which explains that "MVSC initially aimed to develop a solution that could process both renewals and registrations, yet was delayed in its ability to do so because it lacked integration sufficient to meet the needs of the state's point-of-sale requirements." (See [957-5] at 29.) Klein continues: "Given the difficulties of developing a workaround and pressure to get their solution into the market from the state, MVSC eventually decided to bring its renewal solution to market separately. The Interference thus contributed to the delayed timing of MVSC's renewal solution in addition to its registration solution." (Id. at 29-30.) Reynolds fails to address Klein's explanation, leaving a material factual dispute about what caused MVSC's delay in entering Illinois. Reynolds has not established, for purposes of summary judgment, that Illinois's regulatory approval process was an independent, superseding cause of MVSC's delayed entry into the state.

e. Antitrust injury

MVSC must also establish "antitrust injury," which requires proof that its "claimed injuries are of the type the antitrust laws were intended to prevent." Viamedia, 951 F.3d at 481 (quotation and citation omitted). This requirement "'ensures that a plaintiff can recover only if the loss stems from a competition-reducing aspect or effect of the defendant's behavior." Chi. Studio Rental, 940 F.3d at 978 (citation omitted).

Reynolds first claims that MVSC cannot establish antitrust injury because Electron enables MVSC to compete in point-of-sale states. (See Reynolds's MVSC MSJ at 32-33.) Reynolds also argues that harm to a single competitor is not harm to competition, and points to evidence that "throughout the relevant period, competition among EVR providers has been robust, and CVR's competitors ... have either expanded or held their market share while CVR's market share has decreased." (Id. at 33.)

But as MVSC emphasizes, this ignores the expense MVSC undertook in developing Electron. (See Section III.A.3, supra.) It also ignores evidence that MVSC had to delay its entrance into the Illinois and Virginia markets while it developed Electron, resulting in reduced output to consumers. (DRPSOF ¶¶ 75, 116.) In California, MVSC cites evidence that it lost business it would have earned but for its lack of certified or automated access to data. (Id. ¶ 112.) Reviewing the record in the light most favorable to MVSC, it suffered harm from the Reynolds-CDK group-boycott conspiracy because it was unable to compete as effectively. That qualifies as antitrust injury. Sw. Suburban Bd. of Realtors, Inc. v. Beverly Area Planning Ass'n, 830 F.2d 1374, 1382 (7th Cir. 1987).

4. Damages

The court has addressed most of Reynolds's damages arguments. Reynolds primarily emphasizes "MVSC's failure to disaggregate harm caused by an alleged unlawful conspiracy between Reynolds, CDK and/or CVR on the one hand, from harm caused by each defendant's unilateral conduct, conduct by CDK and/or CVR

alone, or state conduct on the other hand." This failure, Reynolds contends, requires summary judgment on the issue of damages. (See Reynolds's MVSC MSJ at 36.) The court already considered and rejected that argument. See Daubert Op., 581 F. Supp. 3d at 1077-81. And the court has likewise considered and rejected Reynolds's argument concerning the disaggregation of damages based on "marketing interference." See Section III. B.1.a, supra.

Reynolds asserts that MVSC cannot claim damages suffered after MVSC settled with CDK in October 2019, and MVSC has not responded to this argument. (See Reynolds's MVSC MSJ at 37; Reynolds MVSC Reply at 17). The court agrees and grants summary judgment in Reynolds's favor on that portion of MVSC's claim. See, e.g., Drug Mart Pharm. Corp. v. Am. Home Prods., Corp., 288 F. Supp. 2d 325, 332-33 (E.D.N.Y. 2003) (settlement agreement is evidence of conspirators' withdrawal from the alleged conspiracy, eliminating post-settlement damages).

C. State Law Claims

1. State antitrust claims

The parties agree that MVSC's Illinois and California state law claims rest on the same theories as the federal antitrust claims and therefore survive summary judgment.

2. California Unfair Competition Law

MVSC concedes that Reynolds is entitled to summary judgment on its claim under California's Unfair Competition Law claim because MVSC was never a Reynolds customer, and Reynolds therefore never took anything from MVSC that could be the subject of a restitution order. See Madrid v. Perot Sys. Corp., 130 Cal. App. 4th 440, 453, 30 Cal.Rptr.3d 210 (2005) ("[I]n the context of the UCL, 'restitution' is limited to the return of property or funds in which the plaintiff has an ownership interest").

CONCLUSION

For the reasons discussed above, Defendants' motions to bar [775, 787] are denied. CDK's motion for summary judgment against AutoLoop [967] is granted as to damages attributable to the Cox Plaintiffs and otherwise denied as to AutoLoop's Sherman Act § 1 conspiracy claim. CDK's motion is granted as to AutoLoop's Sherman Act § 1 exclusive dealing claim and AutoLoop is ordered to show cause why CDK's motion should not also be granted as to AutoLoop's § 2 monopolization claim. CDK's motion on AutoLoop's state law claims is denied to the extent those claims are based on the alleged conspiracy and is otherwise granted.

CDK's motion for summary judgment against the Dealers [970] is granted on all Sherman Act damages claims but denied as to the conspiracy claim for injunctive relief. CDK's motion on the Dealer's state-law claims is granted as to Counts IX, XLIV, XXXIX, XL, XLVI, and XLVIII and is otherwise denied.

Reynolds's motion for summary judgment against MVSC [954] is granted as to any claim based on MVSC's failure to break into the Wisconsin market and as to damages MVSC sustained after settling with CDK in October 2019. Reynolds's motion is also granted as to MVSC's California Unfair Competition Law claim. Reynolds's motion is otherwise denied.