ROVNER, Circuit Judge.

In 2002, Integrated Genomics, Inc. ("IG") granted Tillman Gerngross, a bioengineering professor at Dartmouth College, a license to use genetic sequencing data that it had developed on a common yeast. Gerngross did not disclose to IG that he intended to use the data in connection with a private business venture rather than for purely academic purposes. Gerngross was charged an academic rate for the license which, IG alleges, was much less than it would have charged him for commercial use of the data. Four years later, Gerngross's business was sold for many millions of dollars. IG filed suit against Gerngross, contending that he had defrauded IG by failing to disclose that he intended to use the sequencing data for commercial purposes and that Gerngross had breached his license agreement with IG by "publishing" the licensed sequencing data to his business and to its buyer. The district court entered summary judgment in favor of Gerngross on the contract claim and, after a trial on the fraud claim, found that even if Gerngross was deceitful in failing to disclose his intended use of the data, the evidence did not clearly and convincingly show that IG would have charged him more for the license had it been aware of that use. We affirm.

I.

Gerngross joined the faculty of the Thayer School of Engineering ("Thayer") at Dartmouth in 1998. In 2000, he and a fellow faculty member formed a private corporation, GlycoFi, with the aim of genetically modifying yeasts in such a way that they would manufacture or "express" human proteins with therapeutic and industrial uses. Dartmouth agreed to "incubate" the business, lending its own facilities and personnel to the venture in exchange for an equity stake in the company. Early on, all of GlycoFi's research was performed by a laboratory team at Dartmouth, and all of the supplies and materials used by that team were purchased through the university; the company reimbursed the university for its expenses on a monthly basis. Eventually, GlycoFi would acquire its own facilities and employees, although even then, some work continued at Dartmouth.

Among the species of yeast organisms that Gerngross and his team sought to genetically modify was Pichia pastoris ("Pichia"), a common yeast with characteristics that were amenable to the modification effort. Toward that end, Gerngross sought out genetic sequencing data for that organism. This is IG's business: it maps the genomes of various organisms and sells the data for both commercial and non-commercial, including academic, uses. In the spring of 2002, Gerngross learned that IG was working on the genomic sequence of Pichia and approached the company about acquiring a license of its data. At that time, IG was the sole source from which sequencing data on this organism could be licensed. This is not to say that IG's data was indispensable to GlycoFi's efforts: Gerngross testified that modifying a yeast organism so that it would express human proteins involved introducing various new genes into the organism and eliminating certain existing genes that were not wanted; genomic sequencing data was of no help to the first of these tasks but could be of help to the second. Moreover, according to Gerngross, GlycoFi could have constructed the sequencing data it needed from both limited genetic data on Pichia that was already in the public domain as well as data on similar organisms. But Gerngross believed that obtaining the data from IG might "[s]ave a few weeks here or there." R. 135 at 132. He thus approached IG about obtaining a license for use of its Pichia data.

It is undisputed that Gerngross presented himself to IG's representatives as a professor at Dartmouth, and after hearing conflicting testimony on the question of whether Gerngross disclosed that he was seeking the Pichia data for use by GlycoFi, the district court found that he did not. R. 123 at 4. Gerngross's primary contacts at IG were Dr. Yuri Nikolsky, a co-founder of the company who served as its vice president for business development from 1997 until he left the company in August 2002, and Dr. Yakov Kogan, a research scientist at IG who succeeded Nikolsky as the individual in charge of business development in September 2002 and continued in that capacity until April 2003, when he too left the company. Gerngross testified that he initially spoke with Kogan in the spring of 2002, and when he did so he mentioned to Kogan that he was involved in a commercial venture. But Kogan did not recall speaking with Gerngross at that time; and both he and Nikolsky testified that they were aware of Gerngross' academic affiliation only. As noted, the district court credited IG's witnesses over Gerngross on this point.

If Gerngross did not volunteer his affiliation with GlycoFi, neither did IG inquire. Dr. John Campbell, who assumed responsibility for business development when Kogan departed the company in 2003, testified that many academics are involved with startup commercial ventures, just as Gerngross was. Yet, Nikolsky, who negotiated the Pichia license with Gerngross, could not recall asking him what he planned to do with the Pichia data or whether he intended to use it for commercial purposes. "[A]t the time I was a sales guy, so I didn't ask a whole lot of questions," he admitted. R. 143-4 at 15. To be fair, Campbell testified that, in his experience, dual-affiliation customers were careful to draw lines between their academic and commercial activities, so that there would be no need for a vendor like IG to ask in what capacity the customer was seeking to purchase its data for purposes of establishing a price. Nikolsky at one point in his testimony appeared to contradict Campbell on that point, suggesting that even if an academic were seeking to license data for an industrial use, his academic affiliation alone would determine what IG would charge him for a license. "In our agreement what mattered was affiliation. If the affiliation was a university, that would be considered academic" R. 143-4 at 56; see also id. at 72-73. But at another point in his testimony, Nikolsky seemed to retract that assertion, explaining that in his experience academics typically were not affiliated with industry, and that if IG knew that an academic was working with a commercial venture, he would be charged a higher price. "If you knew that the person is working [for an] industrial company, he would be charged [a] much higher price, yes." R. 143-4 at 66. In any case, given that Gerngross (per the district court's finding) did not disclose his intent to use the data for the benefit of GlycoFi, IG treated him as an academic rather than a commercial customer.

As a general rule, IG charges and has charged commercial licensees more than it does academic licensees for the same data. How much more depends on the circumstances, and IG's witnesses gave a broad and indeterminate range of prices that the company might have charged someone like Gerngross for a commercial license of the Pichia data. Nikolsky, for example, testified that IG gave academics "huge discounts," R. 143-4 at 40, and that industrial customers typically were charged a price that was higher by several times than the fee charged to academics. But Nikolsky could not say what the price would have been for a commercial licensee who, like Gerngross, wanted nothing more than the Pichia data IG already had — i.e., an "off-the-shelf' sale that required no additional work on IG's part. Neither could Kogan. Nikolsky hypothesized that if a commercial customer had sought an exclusive license for particular data — one which would have precluded IG from licensing the same data to anyone else — the price might have run from hundreds of thousands of dollars to over one million; if, on the other hand, the customer wanted a non-exclusive license, as Gerngross did, "[t]hen the price would go down dramatically." R. 143-4 at 57. Dr. Michael Fonstein, IG's co-founder and president until March 2003, testified that he had engaged in negotiations to license the Pichia data commercially to an interested company for a price in the range of half a million to 1.5 million dollars; those negotiations never resulted in a sale, however. In fact, although IG had successfully negotiated commercial licenses on data for other similar organisms, Fonstein was unable to recall any commercial sale of the Pichia data, and perhaps because of that he found it difficult to quantify what the difference in price would have been for a commercial versus an academic license. Campbell similarly testified that although he had engaged in discussions about the sale of an exclusive commercial license of the Pichia data, no such transaction was ever completed. Campbell said that IG did sell a number of non-exclusive commercial licenses on the Pichia data, but he did not name a price for those licenses.

Contributing to the uncertainty as to the price IG might have charged Gerngross in 2002 for a commercial license of the Pichia data is the fact that IG was experiencing significant financial stress in 2001 and 2002. Nikolsky explained that the price for genetic sequencing data was dropping rapidly in those years as a result of technological advances in sequencing and increasing competition in the field. IG, which was one of the larger sequencing firms, began to shed large numbers of employees. Kogan testified that he and some of IG's other remaining employees went unpaid as a result of the company's declining cash flow. Fonstein conceded that the company was in "disarray" and "struggling for survival." R. 143-3 at 16-17. Nikolsky recalled that as a result of the challenging market and IG's precarious financial condition, the prices it charged for its data were fluid. "I think at that time we just needed cash. You know, we needed more sales and just whatever we could get was the price." R. 143-4 at 26.

IG ultimately agreed to provide Gerngross with a copy of its data on the Pichia pastoris genome for $5,000. Nikolsky sent Gerngross a copy of IG's standard academic license agreement for him to sign. The record does not include a copy of that agreement. However, witnesses who were familiar with IG's standard academic agreement (including Fonstein, Kogan, and Campbell) on the whole agreed that it would have restricted the use as well as the publication of the licensed data. Nikolsky was the one exception on this point: he did not recall whether the standard academic agreement would have prohibited commercial use of licensed data.

Gerngross balked at signing that agreement, testifying later that he found its provisions too favorable to IG. By the time of trial, Gerngross could not recall what provisions in particular he found objectionable. At one point in his testimony, Gerngross thought that the draft agreement might have prohibited the commercial use of the Pichia data; at another point, he thought that it might have entitled IG to royalties, which if true would suggest that it might not have been a standard academic agreement at all. In any case, he refused to sign the agreement.

When Nikolsky conveyed to Gerngross that IG's central concern was that the commercial value of the data not be undermined by making it publicly available, Gerngross drafted a letter in which he agreed to restrict his publication of the data. Nikolsky would later testify that a limitation on publication was a key provision of the standard academic agreement. The purpose of this limitation, as Nikolsky had explained to Gerngross, was to protect the commercial value of IG's data, which would drop to zero once the data entered the public domain. Dr. Vinayak Kapatral, who became responsible for IG's business development in 2005, explained in his testimony that an academic who wished to publish his research was typically required to put the sequencing data underlying his or her research in a public database, so that it was available to other researchers for verification purposes. IG's standard academic license thus specified that the licensee could publish a limited amount of the licensed data annually: 10 kilobases (kb). (A kilobase is a unit of measure equal to one thousand base pairs. A base pair comprises two complementary nucleotides or bases which are held together by chemical bonds to form a "rung" of the DNA "ladder"). Kapatral testified that 10 kilobases was roughly equivalent to ten genes' worth of data. IG deemed that quantity sufficient to meet an academic's interest in publication while maintaining the commercial value of its data. (The Pichia genome, for example, contains over nine million base pairs, so 10 kilobases represented a small portion of the genome.)

On April 8, 2002, Gerngross sent a letter to Nikolsky on Thayer letterhead setting forth his agreement to refrain from publishing more than 10 kilobases per year of the licensed Pichia sequencing data:

Dear Dr. Nikolsky

This is to state the restrictions that apply to the data we are obtaining from Integrated Genomics Inc. My research group at Dartmouth College is restricted to the publication of no more than lOkb of sequencing data from the Pichia genome per calendar year. This restriction is void if the genome data becomes available from a public domain at no[] charge.

Sincerely yours, Tillman Gerngross

Joint Ex. 2, R. 142-2 at 2. This is the sole writing in the record which evidences the parties' agreement. Fonstein, among other witnesses, assumed that this must have been part of a broader agreement between the firm and Gerngross. R. 143-3 at 31-32. But there is no documentary evidence of such a broader agreement in the record. Fonstein allowed that some things may have fallen through the cracks during the turbulent period in which Gerngross obtained the license. It "[s]ounds like sloppiness," he remarked. R. 143-3 at 16. Nikolsky likewise could offer no other explanation for the absence of additional documentation.

IG issued an invoice for $5,000 to Dart-mouth and on the following day received a check in that amount from the college. IG then sent to Gerngross at Dartmouth a compact disc containing the Pichia data.

In the ensuing months, GlycoFi began to emerge from the low profile that had characterized its first two years of existence. By the time Gerngross licensed the Pichia data from IG, GlycoFi had obtained a second round of venture capital financing totaling more than seven million dollars. In December, GlycoFi opened its own facility in Lebanon, New Hampshire. Much, although not all, of its ongoing research work moved from Dartmouth to the new facility. The Pichia data, which had been loaded onto the GlycoFi research team's computers at Dartmouth, was moved to the new facility as well.

Although Nikolsky and Kogan would later testify that they regarded Gerngross as an academic user and the license he acquired for the Pichia data to be an academic one, there is evidence that they eventually came to know of his affiliation with GlycoFi. Kogan, for example, admitted that he became aware of that affiliation no later than December 2002, when he sent Gerngross an email in which he stated, "Sometime[] I would like to talk to you as the CSO [Chief Scientific Officer] of . . . GlycoFi. I think that there are some potential joint projects which could benefit both companies." Joint Ex. 5, R. 142-2 at 7. And in January 2003, Gerngross delivered a PowerPoint presentation on "Protein Expression Technology and Applications" to IG at its offices in Chicago. Several IG employees attended, including Kogan and Fonstein. That presentation featured the work Gerngross and his team were doing with Pichia. All but the last of the 49 slides in that presentation displayed the seal of Dartmouth College in the lower left-hand corner and GlycoFi's logo in the lower right. The opening slide of that presentation displayed GlycoFi's logo even more prominently and, in addition, identified Gerngross as both an associate professor at Dartmouth and the CSO of GlycoFi. If anyone at IG was surprised to learn of Gerngross's affiliation with GlycoFi, he or she did not mention it.

Indeed, the former IG personnel who had dealt with Gerngross were notably unwilling to say in retrospect that he had deceived them. Nikolsky declined to say that Gerngross had lied to him or defrauded IG. Kogan did not believe that Gerngross had lied to him or hidden anything from him. And the most that Fonstein was willing to say was that it was "a judgment call" whether Gerngross had misused the Pichia data by putting it to commercial use. R. 143-3 at 46.

IG, according to Gerngross, had orally agreed to provide him with updated Pichia data when it became available, and in April or May of 2003 he telephoned IG to ask for that data. He spoke with Campbell, who was then handling business development for IG. Gerngross identified himself to Campbell as a Dartmouth professor and provided Campbell with his Dartmouth contact information, although by this time GlycoFi had its own facility and Gerngross had a GlycoFi email address. When Campbell checked IG's records, he found no files at all documenting the Gerngross transaction. Yet, he did not inquire of Gerngross about his use of the data, seek to ascertain from Gerngross what restrictions IG had previously placed on his use of the data, or attempt to impose any new or additional restrictions on Gerngross's use of the updated sequence data. Instead, Campbell simply sent to Gerngross at his Dartmouth address a compact disc containing the updated Pichia genomic sequence data. Gerngross understood that the updated data was subject to the same publication restriction as the original data.

In May 2006, the pharmaceutical giant Merck Co. acquired GlycoFi for $400 million, then a record all cash price for the acquisition of a fledgling biotech company. Pursuant to the acquisition, Merck became the owner of all of GlycoFi's assets, including the computer(s) onto which the Pichia data that GlycoFi had licensed from IG had been loaded.

A few months after the acquisition, IG representatives, none of whom were involved with Gerngross's acquisition of the Pichia license in 2002, contacted Merck to express concern that Gerngross was misusing the Pichia data for commercial ends and to demand compensation. When its demands were rebuffed, IG filed suit against Gerngross, alleging that he had fraudulently misrepresented his status in acquiring the Pichia license, breached an alleged oral agreement to use the Pichia data for academic purposes only, and breached his written agreement not to publish more than 10 kilobases of the Pichia genome per year.

The district court granted summary judgment to Gerngross on the two contract claims. Integrated Genomics, Inc. v. Gerngross, No. 07 C 5860, 2009 WL 1940770 (N.D.Ill. July 7, 2009). The court found insufficient evidence of an oral agreement restricting Gerngross's use of the data. Id., at *3. As for the written agreement, the court rejected IG's contention that the sole restriction set forth in that agreement — not to publish more than 10 kilobases of data from the sequence per year — constituted an agreement by Gerngross not to disclose the data to anyone outside of his research group at Dartmouth, including in particular GlycoFi and Merck. The court concluded that the term "publish" should be given its common and ordinary meaning of disclosing information to the public, which Gerngross had not done. Id.

After a trial on the fraudulent misrepresentation claim, the court found no evidence that Gerngross had affirmatively misrepresented the purpose for which he acquired the Pichia data, and no fiduciary obligation on his part that would have required him to volunteer his intention to use it for a commercial purpose. R. 123 at 7-8. IG's best case for fraud, the court reasoned, lay in the possibility that Gerngross had deceived IG by presenting himself as an academic without clarifying that he was seeking the Pichia data for commercial use. R. 123 at 8. Ultimately, however, IG had not convincingly shown that the clarification would have made a difference in the price it charged Gerngross for the data:

A solid argument can be made that when Gerngross represented himself as a Dartmouth professor and advised Integrated Genomics that the data would be used by his research group at Dartmouth, those representations were technically true but misleading. Integrated Genomics has failed to show by clear and convincing evidence, however, that the undisclosed information was material or that it would have acted differently had it known of Gerngross's commercial affiliation Integrated Genomics did not establish that it had a hard-and-fast practice at the time of charging more to dual-affiliation customers like Gerngross or that it would have charged him more had it been aware that the data would be used in GlycoFi's work. In addition, there is no evidence that Integrated Genomics asked Gerngross how he intended to use the Pichia data. The evidence reflects that at the time in question, Integrated Genomics was not particularly careful about such matters, which is consistent with its failure to inquire into Gerngross's purpose. Indeed, the limited restriction that Integrated Genomics imposed when it sold the Pichia data to Gerngross did not restrict him from commercial use of the data. This indicates that the possibility that an academically-affiliated customer might also use data for commercial purposes was not a particularly significant factor for Integrated Genomics at the time.

R. 123 at 8-9.

IG appeals the district court's adverse rulings on the claim that Gerngross breached his written agreement not to publish more than ten kilobases of the licensed Pichia data per year and on the claim that Gerngross fraudulently misrepresented his status in obtaining the license from IG.

II.

A. Breach of Contract — "Publication"

We begin with IG's claim for breach of a contract. The sole written agreement between Gerngross and IG is found in Gerngross's April 8, 2002 letter to IG's Nikolsky, in which Gerngross acknowledged the following limitation on disclosure of the Pichia data: "My research group at Dartmouth College is restricted to the publication of no more than 10kb of sequencing data from the Pichia genome per calendar year." Joint Ex. 2, R. 142-2 at 2. IG's contention is that Gerngross breached that limitation when he shared the entirety of the Pichia data with GlycoFi and in turn with Merck. The key question, then, is whether communication of the data to GlycoFi and Merck constituted "publication" of the data as that term was used in the agreement. The district court, of course, concluded that it did not, reasoning that "publication" ordinarily connotes disclosure to the public. As this claim was disposed of on summary judgment, our review is de novo. E.g., Normarir-Nunnery v. Madison Area Tech. Coll., 625 F.3d 422, 428 (7th Cir. 2010).

In this diversity action, state law governs the substance of IG's claim. Erie R.R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938). The parties assume without discussion that Illinois law governs the contract between Gerngross and IG, a corporation whose principal place of business is in Illinois, and we have no reason to question that assumption. Under Illinois law, "[c]ourts interpret contracts with the goal of effectuating the parties' intent, giving contract terms their plain and ordinary meaning." Kim v. Carter's Inc., 598 F.3d 362, 364 (7th Cir. 2010) (citing Hot Light Brands, LLC v. Harris Realty Inc., 392 Ill.App.3d 493, 332 Ill.Dec. 72, 912 N.E.2d 258, 263 (2009)). Of course, the pertinent language must be viewed in context, and the contract must be construed not in a piecemeal fashion but as a whole in determining the parties' intent. Utility Audit, Inc. v. Horace Mann Serv. Corp., 383 F.3d 683, 687 (7th Cir. 2004) (citing Trade Center, Inc. v. Dominick's Finer Foods, Inc., 304 111. App.3d 931, 238 IlLDec. 230, 711 N.E.2d 333, 335 (1999)).

We believe that the district court correctly understood the plain and ordinary meaning of the term "publication" to signify disclosure to the public, rather than the disclosure of information to another individual or corporation within the context of a business or professional relationship. See BLACK'S LAW DICTIONARY 1242 (7th ed. 1999) ("Generally, the act of declaring or announcing to the public"); OXFORD ENGLISH DICTIONARY ONLINE, http://english. oxforddictionaries.com (last visited Feb. 21, 2011) ("the action of making something generally known"); MERRIAM-WEBSTER DICTIONARY ONLINE, http.7/merriam-webster.com/dictionary ("the act or process of publishing"; in turn defining publish as "to make generally known," "to make public announcement of and "to disseminate to the public") (last visited Feb. 21, 2011); DICTIONARY.COM, http:// dictionary.reference.com ("the act of bringing before the public; announcement") (last visited Feb. 21, 2011). This understanding is consistent with the term's derivation from the Latin verb publieare, see OXFORD ENGLISH DICTIONARY ONLINE, http://english.oxforddictionaries.com, which means "to appropriate to the public use," CASSELL'S LATIN DICTIONARY 486 (1977). Gerngross never disclosed the Pichia data to the public, but he did share the information with both GlycoFi and Merck; and IG contends that "publication" should be understood to include the communication of information to an individual, or in this case, a corporation, as well as the public at large. But no common understanding of the term "publication" includes such a limited disclosure of information.

IG relies on the Illinois Supreme Court's decision in Valley Forge Ins. Co. v. Swiderski Elecs., Inc., 223 Ill.2d 352, 307 111. Dec. 653, 860 N.E.2d 307 (2006), for the proposition that Illinois law defines "publication" to include the simple communication of information, whether it be to one person or to many. But Valley Forge is of little help to IG. The court in Valley Forge was called on to interpret the term "publication" in the context of certain insurance policy provisions obligating an insurer to defend its insured against suits for damages caused by "personal and advertising injury." The policies in question defined "personal and advertising injury" to include "[o]ral or written publication, in any manner, of material that violates a person's right to privacy" and, similarly, "[o]ral, written, televised or videotaped publication of material that violates a person's right of privacy." Id., 307 IlLDec. 653, 860 N.E.2d at 311. Swiderski Electronics, the insured, was sued for faxing unsolicited advertisements to numerous persons and businesses in violation of Telephone Consumer Protection Act, 47 U.S.C. § 227. The question before the court was whether the faxes constituted the publication of material that intruded upon one's right to privacy, such that the insurance companies had a duty to supply Swiderski with a defense. The terms "publication", "material," and "privacy" were left undefined by the policies. The court accorded "publication" its plain and ordinary meaning of communicating information to the public, id., 307 Ill.Dec. 653, 860 N.E.2d at 316-17, and reasoned that "[b]y faxing advertisements to the proposed class of fax recipients as alleged in [the] complaint, Swiderski published the advertisements both in the general sense of communicating information to the public and in the sense of distributing copies of the advertisements to the public," id., 307 IlLDec. 653, 860 N.E.2d at 317. Based on its construction of the term "publication," as well as the other policy terms, the court concluded that Swiderski's unsolicited faxes could be found to constitute written publications that violated the recipients' right to privacy. The suit against Swiderski was therefore properly characterized as one seeking compensation for "personal and advertising injury," and the insurers were required to supply Swiderski with a defense. Insofar as it is relevant here, the Valley Forge decision adopted the same meaning of the term "publication" — the communication of information to the public — that we have.

It is true, as IG points out, that we later described Valley Forge as "interpret[ing] `publication' to mean nothing more than `communication.`" Auto-Oimers Ins. Co. v. Websolv Computing, Inc., 580 F.3d 543, 550 (7th Cir. 2009), cert. denied, ___ U.S.___, 130 S.Ct. 1884, 176 L.Ed.2d 363 (2010). But the point we were making in Auto-Owners was about the type of publication that interferes with one's right to privacy. Recall that the insurance policies at issue in Valley Forge covered suits for "advertising and personal injury," which they defined in relevant part as the publication of material that interferes with a person's privacy right. This court, among others, had held that in order to violate that right, the publication must disclose some secret or personal information. Am. States Ins. Co. v. Capital Assocs. of Jackson County, Inc., 392 F.3d 939, 942-43 (7th Cir. 2004). By contrast, Valley Forge reasoned that a publication in the form of an unsolicited fax advertisement violates the recipient's right to privacy in the sense that it intrudes upon his seclusion. 307 Ill.Dec. 653, 860 N.E.2d at 317-18. It was this divergence of understanding as to the confidential nature of the information that is published that we were addressing when we said that the Illinois Supreme Court had defined "publication" to mean nothing more than "communication." Auto-Owners, 580 F.3d at 551. We were not construing Valley Forge to hold that under Illinois law, any means by which information is communicated constitutes publication, even if the information is communicated privately to one's business associate as opposed to the public. The Valley Forge decision itself makes clear that the fax communication at issue there was with members of the public at large.

B. Fraudulent Misrepresentation — Gerngross's Academic Affiliation

IG's claim of fraudulent misrepresentation is premised on the notion that Gerngross presented himself to IG as an academic and told its employees that the Pichia data would be used by his research group at Dartmouth, without disclosing that the work he and his research group were doing was for the benefit of a commercial venture (GlycoFi) then being incubated at Dartmouth. Had it realized that Gerngross intended to use the data for commercial purposes, IG maintains, it would have charged him much more for the data; and that is precisely why, IG asserts, that Gerngross kept his affiliation with GlycoFi to himself. In order to prevail on its claim, IG bore the burden of proving: (1) Gerngross made a false statement of material fact; (2) that he did so knowing or believing the statement to be false; (3) that he made the statement with the intent to induce IG to act; (4) that IG took action in justifiable reliance on the truth of the statement; and (5) IG suffered damage as a result of its reliance on the statement. See, e.g., Doe v. Dilling, 228 Ill.2d 324, 320 Ill.Dec. 807, 888 N.E.2d 24, 35-36 (2008). IG agrees that it was obligated to prove each of the elements of its fraud claim by clear and convincing evidence. See Avery v. State Farm Mut. Auto. Ins. Co., 216 Ill.2d 100, 296 IlLDec. 448, 835 N.E.2d 801, 856 (2005); Ass'n Ben. Servs., Inc. v. Caremark RX, Inc., 493 F.3d 841, 853-54 (7th Cir. 2007); Barrington Press, Inc. v. Morey, 752 F.2d 307, 309-10 (7th Cir. 1985). Although there is no evidence that Gerngross made any affirmatively false statement to an IG representative, a false statement may include a half-truth which, although technically accurate, is misleading because it omits important qualifying information that, had it been known, would have caused the plaintiff to act differently. Williams v. Chicago Osteopathic Health Sys., 274 Ill.App.3d 1039, 211 Ill.Dec. 151, 654 N.E.2d 613, 622 (1995) (citing Lindsey v. Edgar, 129 111. App.3d 718, 84 Ill.Dec. 876, 473 N.E.2d 92, 95-96 (1984), and Huls v. Clifton, Gunderson Co., 179 Ill.App.3d 904, 128 IlLDec. 858, 535 N.E.2d 72, 76 (1989)). Gerngross's presentation of himself to IG as an academic is thus claimed to be fraudulent because he omitted the qualifying information that the purpose for which he was seeking the Pichia data was a commercial rather than an academic purpose. The district court, as we have noted, agreed that there was "[a] solid argument" for the proposition that Gerngross's failure to disclose his commercial intentions was misleading, but found that IG had not proved by clear and convincing evidence that the omission was material or, relatedly, that it would have acted differently had it been aware of the omitted information. R. 123 at 8. The court's determination that there was no fraud, as an application of a legal standard to the particular facts developed at trial, is a determination that we review for clear error. See SEC v. Maio, 51 F.3d 623, 636 n. 16 (7th Cir. 1995) (citing Ambrosino v. Rodman Renshaw, Inc., 972 F.2d 776, 785 (7th Cir. 1992)); Barrington Press, 752 F.2d at 310.

We cannot say that the district court's determination was clearly erroneous. Certainly there is evidence to support the conclusion that IG, as a general rule, charges more to commercial than academic users of its data. And there is ample evidence that Gerngross in his dealings with IG emphasized his academic connections while remaining silent about GlycoFi — even after GlycoFi had begun to emerge from its incubation at Dartmouth, as it had done by the spring of 2003, when Gerngross sought the update of the Pichia data from IG. It is entirely possible that a different factfinder might have concluded that Gerngross's failure to disclose the commercial purpose for which he sought the Pichia data was material, and that IG likely would have charged Gerngross a substantially higher price for the data had he been more forthcoming. But the evidence also supports the contrary conclusion, and it thus leaves ample room for the district court's conclusion that fraud had not been proven.

First, IG did not require Gerngross to sign an agreement precluding him from making commercial use of the Pichia data. As we have noted, the record indicates that IG typically did include such a use restriction in its agreements with academic licensees: Kogan, Fonstein, and Campbell all identified that as a standard provision in academic licenses, although Nikolsky was uncertain on that point. Whether there was such a provision in the proposed contract that IG gave to Gerngross, and that he refused to sign, is unknown. Neither party produced a copy of that contract, Nikolsky could recall little about the details of his negotiations with Gerngross or the terms of the proposed agreement, and Gerngross likewise could remember nothing specific about the provisions other than that he found them on the whole to be too favorable to IG. But regardless of why Gerngross said he would not sign the proposed contract, the record is devoid of evidence that Gerngross agreed to any restriction on his use of the Pichia data. The letter that Gerngross prepared and sent to Nikolsky is the sole evidence of any agreement between Gerngross and IG, and that letter sets forth a single provision which restricted only the amount of data that Gerngross could publish, nothing more. The fact that IG was willing to give up the commercial use restriction it typically imposed on academic researchers is, at the least, consistent with the notion that IG was either aware of or indifferent to the possibility that Gerngross had commercial plans for the data.

Second, IG's interactions with Gerngross do not reflect particular care on the part of IG employees in ascertaining the purpose for which he was seeking the Pichia data. Campbell agreed that it was common for prospective licensees to have dual affiliations with both academia and private enterprise. Even if individuals with business as well as academic ties typically were careful to separate the two in their dealings with others, as Campbell averred, one would think that a company which typically charged more for the commercial as opposed to the academic use of its data would take the initiative in ascertaining a prospective licensee's plans for the data. Yet, Nikolsky agreed that he did not ask many questions of Gerngross as to his reasons for obtaining a license to use the Pichia data. Campbell himself turned over an updated version of the data to Gerngross in 2003 without inquiry, even after he could not find any documentation of IG's agreement with Gerngross in the company's files. IG was on notice of Gerngross's commercial affiliation at least by January of that year (before Campbell turned over the update) when Gerngross made the PowerPoint presentation to IG employees which displayed his affiliation with GlycoFi as well as Dartmouth. That presentation focused on Gerngross's work with the very Pichia organism whose genetic data Gerngross had licensed from IG. But even then, no one at IG took the trouble to inquire or clarify in what capacity Gerngross was using the data. The company's lack of curiosity about Gerngross's use of the data supports the notion that the use did not matter in terms of the price charged for the license.

Third, Gerngross was seeking neither an exclusive license of the Pichia data, which would have prohibited IG from licensing the same data to other customers, nor any customized work vis-a-vis that data on the part of IG. These were both factors that the witnesses identified as reasons why IG might charge a significantly higher price for a license. Gerngross was seeking a non-exclusive license for data that IG had already prepared and could provide to him with no further work on IG's part (other than to supply him with a subsequent update of whatever additional data on the Pichia genome it had assembled). Nikolsky and other witnesses agreed that the price for a non-exclusive license of "off-the-shelf" data would be significantly lower; and Nikolsky testified that the $5,000 price that Gerngross paid was not out of the ordinary for data of this kind.

Indeed, neither Kogan nor Nikolsky, the two IG employees with whom Gerngross dealt in 2002, was willing to say in retrospect that he believed himself to have been deceived by Gerngross. It is true that both Kogan and Nikolsky were long gone from IG by the time they were deposed in this litigation. Nonetheless, they were the company's principal contacts with Gerngross, and as such their testimony is illuminating. If, as IG now maintains, the company invariably charged its clients a much higher price for a license that permitted commercial use of its data, then one would expect Kogan and Nikolsky to have testified that, yes, had they known that Gerngross was seeking the data for GlycoFi, they would have negotiated a higher fee for the license.

Finally, the evidence supporting the notion that IG would have charged Gerngross a substantially greater fee for a commercial license is inconclusive. The testimony does establish subsequent efforts by IG to license the Pichia data for commercial use at much, much higher prices than IG charged Gerngross, although none of these negotiations actually came to fruition. Furthermore, IG did finalize other licensing agreements at higher prices with commercial users vis-à-vis genetic sequencing data on organisms other than Pichia. Collectively, these efforts to charge higher prices to commercial users, both successful and unsuccessful, certainly support IG's contention that it would have charged Gerngross more had it understood the capacity in which he was seeking the Pichia license. On the other hand, IG had been undergoing significant financial distress when Gerngross sought the Pichia license in 2002. Kogan testified that the company could not meet its payroll, Fonstein said that the company was "struggling for survival," R. 143-3 at 16-17, and Nikolsky said that the company's need for cash was such that "just whatever we could get was the price," R. 143-4 at 26. It is a fair inference from this testimony that even if Gerngross had been more forthcoming about his affiliation with GlycoFi and his intent to use the Pichia data for commercial purposes, IG would have been eager to close the sale and obtain some much needed cash by giving him the academic rate. This may explain why neither Nikolsky nor Kogan was willing to testify that he was deceived by Gerngross or that he would have done something differently had he known what hat Gerngross was wearing in obtaining the Pichia license from IG. It may also explain why IG did not take further steps to verify the purposes for which Gerngross sought the data and/or to affirmatively restrict his use of the data. In any case, the accounts of IG's financial distress in 2001 and 2002 permit the conclusion that IG would not have charged Gerngross more for the data had it realized he meant to use the data for commercial rather than academic purposes.

III.

IG had the burden to establish that Gerngross committed fraud by clear and convincing evidence. Among other things, IG was obliged to show that the intended use of the data it licensed to Gerngross was material to the price charged and that the company would have acted differently (i.e., charged him more) had it realized he intended to use the data for commercial rather than academic purposes. Although the record may have supported a finding that IG certainly would have charged Gerngross more for the Pichia data had it known that he intended to use the data in furtherance of GlycoFi's commercial aims, the evidence also supports the contrary inference. The job of resolving such competing inferences falls to the factfinder. IG's appeal invites us to weigh the evidence differently than the district judge did, which is not within our province. The record reveals that the district judge heard, understood, and evaluated the evidence conscientiously. The judge concluded that IG failed to present clear and convincing proof that IG would have charged Gerngross a higher price even if he had been more forthcoming about his plans, and given the mix of evidence presented, there was nothing erroneous about that conclusion. The district judge also correctly construed the term "publication" in Gerngross's agreement with IG to limit only his public disclosure of the Pichia data rather than his ability to share the data with his company or its purchaser. For these reasons, we AFFIRM the judgment in favor of Gerngross.