Summary
finding a single product where the alleged tie was the result of a "new technique" added to the defendant's existing data-gathering product in an "attempt to improve" it and reasoning that the product "cannot be broken down into each miniscule type of analysis but rather must be looked at as an overall service"
Summary of this case from Siva v. Am. Bd. of RadiologyOpinion
No. 84 C 6704.
November 28, 1984.
Thomas W. Johnston, Steven C. Page, Michael L. Ile, Champ W. Davis, Jr., Chadwell Kayser, Chicago, Ill., for plaintiff.
Thomas F. Ryan, Christian L. Campbell, Chaim T. Kiffel, Sidley Austin, Chicago, Ill., James W. Carter, Jr., A.C. Nielsen Co., Northbrook, Ill., for defendant.
MEMORANDUM OPINION AND ORDER
Marketing research service is the furnishing of usable statistical information to the manufacturers of various goods about the movement of consumer goods through the retailers to the consumer. The raw data from which this information is processed is obtained by counting. Counting here involves the use of sampling traditional to the concept of survey research. Counting can be done at several points along the path taken by the goods from the manufacturer to the consumer; at the manufacturer's shipping dock, at the entrance and exit docks of the distributors or warehousemen, at the entrance doors and in the store rooms of the retailer, at the place of the shelving or deshelving of the goods in the store of the retailer, at the check-out counter of the retailer, and indeed, at the shelving or deshelving of the goods in the consumer's house.
The statistics that can be developed from the raw data, and the uses to which they can be put, vary in meaning, in reliability and in usefulness to the producer, both in his production planning and his advertising planning, according to the point or combination of points at which the raw data is gathered, the frequency with which it is gathered, and the selectivity of the sampling used in the system.
The defendant in this case, one of the old and experienced firms in the field of sampling survey research has devoted one of its departments to the selling of marketing research services to consumer goods manufacturers (principally manufacturers in the food industry) for over thirty years. With a wide spectrum sampling base, its research statistics were relatively predictable nationally. It concentrated on using raw material gathered from storage room and shelf counts in the retailer's stores. It did not use the manufacturer or warehousemen's storage racks or docks for raw data, and it did not use data gathered from the consumer. For purposes of counting it used human beings; its own employees going into the stores, counting from their shelves and store rooms, equipped with pencils and pads. Its product delivered to its manufacturing customer has been a substantial bi-monthly publication and interim advisories called the Nielsen Food Index.
Recently, an innovation called scanning developed in the grocery store business, and particularly in the chain supermarkets for helping to speed along the long lines at the checkout counters, and for improving accuracy in cash register totalling. Now there has developed among manufacturers universal product codes which they place on packaged products at the time their products are packaged. Then at the checkout counter these codes are picked up by an electric reading lamp when the package of goods are scraped across a glass screen. The process is called scanning.
Scanning has revolutionized the whole field of retail-exit counting and is particularly adaptable to the field of marketing research of consumer goods. Now scanners not only can make computer read-outs for the store itself but the information can be transferred to the computers of businesses engaged in marketing research of consumer goods.
The plaintiff in this case was founded seven years ago for the express purpose of selling a research service to manufacturers, using the process of scanning. It has several research products, the most important of which are BehaviorScan and the Marketing Fact Book. Its founder and Chief Executive Officer described his business in these words:
"IRI began selling BehaviorScan its first product, in 1979. BehaviorScan is a test marketing service. IRI has installed scanning equipment which is able to read the Universal Product Codes (UPC) on packaged grocery products at the checkout facilities in supermarkets representing approximately 90% of the total supermarket sales volume in each of eight markets. IRI also has assembled consumer panels consisting of approximately 2,500 households in each market. By using this scanning equipment and IRI's computerized data collection system, IRI gathers household-by-household data on the purchasing behavior of its consumer panels, as well as data on the total sales of all UPC coded packaged goods made by the participating supermarkets. These data are used to analyze consumer response to new products, advertising, pricing strategies and other marketing techniques employed by manufacturers and distributors of packaged consumer goods.
"Each of IRI's eight markets has cable television. IRI is able to direct alternate television advertising to its consumer panel households. Together with IRI's scanning data collection capability, this permits the measurement of direct consumer response to television advertising. IRI also administers, supervises and performs product test marketing and various other merchandising activities on behalf of its clients in each of the eight markets.
"IRI developed and in 1981 began selling its second major product, the Marketing Fact Book (MFB). The MFB constitutes a separate and distinct application of IRI's scanning data base. Whereas BehaviorScan is essentially a test marketing service, the MFB tracks weekly sales results together with reports on ad features, store displays, shelf price reductions, coupons and other "causal" marketing information. These data enable packaged goods manufacturers to measure the effect of price changes and promotions on sales. The MFB data base also contains demographic data which permit manufacturers to determine the types of customers who are buying their products.
"The MFB is issued in hard copy quarterly and annually, but its real value lies in the fact the the MFB data base is directly accessible by computer. IRI's clients can therefore use the vast MFB data base to perform frequent and timely analyses of product sales performance and measure the effect of price changes and promotions on sales."
There are several other market research service companies which have been coming into this business during the last few years, using one type or another data base in an effort to produce a distinctive service product, but none of them including the plaintiff here has developed anything near the size of the defendant. One of them appeared by witness to testify on behalf of the plaintiff in this first case testing the products of marketing research in the scanner computer age. Several years ago, even Nielsen began experimenting with the use of scanner data in place of manual audit procedures, and set up two programs; one called Local Markets and one called National Markets.
What ignited this case, which, incidentally began with a Petition for a Temporary Restraining Order that enlarged to a Petition for a Preliminary Injunction, was the announcement in June of 1984 by Nielsen of its intention to dive full force into the use of scanners and scanner data while retaining as long as necessary its now antiquated system of manual audit, providing as the system developed more custom tailored research information for the manufacturer, standardizing research information output on a two week or semi-monthly basis instead of a bi-monthly basis, and providing in due course of time feed back information directly to the retailer. The defendant calls the program which was scheduled to get underway on November 1, 1984, an Enhanced Nielsen Food Index, and outlines what it calls a Transition Period of two to four years, beginning on November 1st.
Because of the imminence of this so called Transition Period, plaintiff has asked this court to issue a preliminary injunction forthwith, basing its claim on the "tying concept" of the Anti-Trust Laws. Plaintiff says that defendant is tying its new product to the old product over which it has exercised extensive economic power for many years and that this action is in Restraint of Competition, conduct proscribed by Section I of the Sherman Act. The language plaintiff seeks in the injunction describes distinctly the nature of this case:
". . . that Nielsen, its officers, agents, servants, employees and persons acting under its direction and authority or in concert with it are hereby enjoined and restrained from selling Nielsen Food Index (NFI) audit data, information generated from such data and/or consulting services relating thereto . . . on condition that purchasers also buy from Nielsen scanner data, information generated from scanner data and/or consulting services relating thereto. Nothing herein shall prevent Nielsen from selling scanner data separate and apart from audit data."
The key to this case is the answer in each instance to the question "when in this new and still developing industry is a product a new one or merely an improvement of an older one?" This question must first be answered because of the defendant's Motion to Dismiss based on the ground that the plaintiff lacks standing to sue; i.e., that the plaintiff is not in competition with the defendant. Defendant urges that plaintiff's product deals with consumer behavior whereas defendant's product deals exclusively with the movement of grocery items across the facilities of the retailer store. As has been stated before, the products we are dealing with are not the raw data nor the method of collecting it. The products are refined residuum of such raw data, refined by computer and programmed analysis, with reasoned explanations and projections made available periodically to the manufacturers of the food items we all of us eventually purchase in the stores.
It appears from the evidence that some part of plaintiff's distributed product includes information that is unique because plaintiff is in a position to translate for its customers facts about the purchasing and consuming practices and habits of consumers and their behavior patterns that can be processed or inferred from sales data supplied by consumers who agree to serve as "guinea pigs" for the plaintiff. At the same time, plaintiff is engaged in gathering and processing data from the stores themselves.
The statutory provision relevant to the issue of standing is section 16 of the Clayton Act, 15 U.S.C. § 26. To establish standing under section 16, the plaintiff must show a "threatened loss or damage by a violation of the antitrust laws." Unless the products are competitive or potentially competitive, there can be no threatened loss or damage. In this relatively new and rapidly expanding industry, the concept of the product must not be drawn too narrowly. Simply because one company might measure only eight markets and another might measure hundreds of markets does not mean that the two companies do not compete now or will not compete in the future as the smaller company grows. Similarly, just because one company uses its raw data to provide a certain type of marketing information, say consumer behavior information, and another company uses its raw data to provide a different type of marketing information, say market share information, does not mean that those companies would not be potential competitors as their products grow and change with advances in technology. The concept of the product here must be painted with a broad brush. Both IRI and A.C. Nielsen provide information to manufacturers on product movement. The details of market share, market size, consumer profiles, etc. are merely small components of an overall picture given to a manufacturer to assess how its product is being received in the market place. To draw the line based on these details seems to be asking too much. Therefore, I find that on the issue of standing there is potential competition. Even at this time there is some actual competitive overlap in the area of pricing and promotional analysis, even though the analysis in those areas is reached by different procedures. As the products and data gathering techniques advance technologically there will continue to be potential competition in other areas as well. For these reasons, the case cannot be dismissed for lack of standing to sue.
I turn now to the plaintiff's request for a preliminary injunction. When a district court is asked to issue a preliminary injunction it should consider four factors: (1) whether the plaintiff will have an adequate remedy at law or will be irreparably harmed if the injunction does not issue; (2) whether the threatened injury to the plaintiff outweighs the threatened harm the injunction may inflict on the defendant; (3) whether the plaintiff has a reasonable likelihood of success on the merits; and (4) whether the issuance of the injunction will serve the public interest. Libertarian Party v. Packard, 741 F.2d 981, 984 (7th Cir. 1984).
The first factor boils down to whether damages can be ascertained in monetary terms. I am unconvinced by plaintiff's assertion that damages would be unreasonably difficult to calculate. If necessary IRI can survey any customer which leaves them and see if they went to A.C. Nielsen for marketing research analysis. Alternatively, IRI could poll every new customer or even each existing customer of Nielsen and see if these customers would have used the services of IRI were it not for the enhancement in Nielsen's product. In short, damages are ascertainable.
The second factor also favors Nielsen. If a preliminary injunction were to issue at this time, it would cause Nielsen's $25 million in improvements to be held in abeyance. Customers would have to be apprised of the injunction and contracts would have to be halted or rewritten. If on the other hand, an injunction were not issued and this were later found to have been erroneous, the harm to IRI could be calculated in damages and any long-term contracts between Nielsen and its customers could be halted at that time. Thus, the injury to the defendant, if an injunction were to issue, would outweigh the injury to the plaintiff if such an injunction were not issued.
The third factor is the critical factor in this case. The plaintiff must prove a reasonable likelihood of success on the merits. In order to prove this likelihood of success, the plaintiff must show the court that there is a reasonable probability that the elements of a tying violation would be established at a trial on the merits. The elements of a tying violation which the plaintiff must show are: (1) two separate products or services; (2) substantial market power in the tying product; (3) use of Nielsen's market power in the tying product to coerce sales of the tied product and; (4) foreclosure of sales by competitors in the tied product market. Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2, 104 S.Ct. 1551, 1556-56, 80 L.Ed.2d 2 (1984).
The basis on which all four factors must rest is whether there are two separate products, for without a "tying" and a "tied" product the last three factors are meaningless. It is this single factor on which plaintiffs have not shown a convincing case at this time. A.C. Nielsen provides its customers with marketing information. The data gathering method it has used for years has involved manual audits. Now, to take advantage of advanced technology, it is using scanners to gather the same data. This advance in technology will allow the data to be gathered more quickly and more economically. Admittedly, because of the increase in speed with which Nielsen is able to receive the data, other types of marketing analysis will be possible. The product must be viewed broadly, as stated earlier. It cannot be broken down into each miniscule type of analysis but rather must be looked at as an overall service. This broad view of the product which allows the plaintiff to have standing works against the plaintiff on the question of whether there are two separate products. This new technique of data gathering is a product improvement. The data generated by scanners and the data generated by manual audits may be used separately and in conjunction with each other during this time of transition.
"The practice has been to classify a product as a single product if there are obvious economies of joint provision, as in the left-shoe-right-shoe example." Jack Walters Sons v. Morton Building Inc., 737 F.2d 698, 703 (7th Cir. 1984).
Here, there are obvious economies of joint provision during this time.
"Very few products today are produced like a statue that is made by pouring molten bronze into a cast — the ultimate single product. But to hold therefore that every composite product is a tie-in, subject to the hostile scrutiny to which antitrust law still subjects tie-ins, would place industry under a vast antitrust cloud and has been rejected." Id.
A different approach to the question of separate products was taken in the Jefferson Parish case where the Supreme Court held that products are separate for tie-in purposes if there are separate markets for each product. 104 S.Ct. at 1562-63. Here, Nielsen has been losing NFI customers apparently because of the outmoded method of data gathering which results in a greater delay from the time of data gathering to the time of the release of the marketing information. It is to keep up with the customer demands for the use of the new technology that Nielsen is attempting to improve its product in this way. Customer demand, whether overt or implied by loss of customers to a scanner-based service such as SAMI, shows that it is the same market demanding the improved service.
"The answer to the question whether petitioners have utilized a tying arrangement must be based on whether there is a possibility that the economic effect of the arrangement is that condemned by the rule against tying — that petitioners have foreclosed competition on the merits in a product market distinct from the market for the tying item." Jefferson Parish v. Hyde, 466 U.S. 2, 104 S.Ct. 1551, 1563, 80 L.Ed.2d 2 (1984).
Here, the markets are not clearly distinct. Realistically, the product must not be reduced to its smallest elements merely to create separate markets.
In sum, the plaintiff has not yet shown a convincing case that there is a reasonable likelihood of success on the merits because it has not been able to show that the enhanced NFI is a product or service separate from the NFI as it had existed through the use of manual audits only.
I turn now to the final factor to be considered in the issuance of a preliminary injunction: whether the issuance of such an injunction will serve the public interest. IRI is currently the leader in scanning-generated market research information. IRI is expanding into the area of markets in larger cities. The presence of the defendant already in these areas tends to increase competition and that is consistent with the purposes of the antitrust laws. For these reasons and because it appears unlikely that a tying case can be made out at this time, I find that the issuance of a preliminary injunction would not serve the public interest.
The plaintiff's request for a preliminary injunction is denied at this time. The parties may choose to request the court to certify this case before proceeding with further discovery and preparation for trial.