DORR-OLIVER, INC. v. FLUID-QUIP, INC.

FLAUM, Circuit Judge.

Plaintiff Dorr-Oliver, Inc. ("Dorr-Oliver") filed suit in district court, claiming that defendants Fluid-Quip, Inc. ("Fluid-Quip"), and its president and majority stockholder, Andrew Franko, copied the trade dress of an industrial machine sold by Dorr-Oliver, in violation of the Lanham Act, 15 U.S.C. § 1051, et seq., and various state law provisions. After a bench trial, the district court ruled that Dorr-Oliver had established the infringement of its trade dress. Consequently, the court enjoined Fluid-Quip from selling its infringing product and awarded Dorr-Oliver monetary relief representing Fluid-Quip's profits from sales of the infringing machine. The defendants appeal the district court's decision on a number of grounds. We hold that Dorr-Oliver failed to establish a likelihood of consumer confusion and therefore reverse the judgment of the district court.

I.

The dispute in the current case requires us to take a sojourn into the field of the corn wet milling industry. Corn wet milling plants process corn by separating the four elements of the corn kernel — germ, fiber, starch, and protein. The separated elements are then sold and used to make food products, such as corn oil and corn starch. In the 1950's Dorr-Oliver invented a new method for separating starch from protein through the use of centrifugal force. This "starch washing" process employs a series of machines, known as "starch washers," to accomplish the goal of separation. The mixture of starch and protein is pumped through the starch washers under high pressure and is distributed through hundreds of small, stationary, tapered, plastic tubes called "cyclonettes." Centrifugal force is created in the cyclonettes, causing the starch and protein to exit through opposite ends of the cyclonettes and into separate outlet ports of the starch washer.

Dorr-Oliver manufactures and sells different kinds of starch washers, which vary in their outward appearance. Dorr-Oliver's claim of trade dress infringement is based on the external design of one type of starch washer, known as the "DorrClone Type C" or the "clamshell." The clamshell was developed and introduced by Dorr-Oliver in the late 1950's under its registered "DorrClone" trademark. The clamshell housing is made of cast steel and, in the words of the district court, resembles a forty-inch diameter bagel with a cylindrical core plugged into its center.1 Various pipes protrude from the "bagel," which is supported vertically above the ground by two legs.2 Each clamshell sold by Dorr-Oliver has a nameplate displaying the "Dorr-Oliver" name, the DorrClone trademark, and the numbers of several expired patents relating to the machine. The name "Dorr-Oliver" is also cast into both sides of the clamshell's outer housing. The other Dorr-Oliver starch washer models bear no resemblance to the clamshell.

The market for clamshell starch washers is very limited. In the United States, there are only twelve purchasers of clamshells for twenty-seven corn wet milling plants. All twelve businesses in the clamshell-purchasing market own and operate Dorr-Oliver clamshells. The clamshells need only be purchased once because their steel outer-housings last indefinitely. The plastic cyclonettes contained in the clamshells, however, periodically need to be replaced. In 1989 Fluid-Quip began selling replacement parts for various kinds of equipment used in the corn wet milling industry. Yet for many years, long after its patents on the clamshell had expired, Dorr-Oliver remained the sole producer and supplier of the clamshell. Dorr-Oliver's customers apparently tired of paying monopoly prices for the clamshells. In 1991, Fluid-Quip was approached by several of Dorr-Oliver's customers, who asked whether Fluid-Quip could produce and supply clamshells at a more reasonable price. The customers desired that Fluid-Quip's clamshells be completely interchangeable with those of Dorr-Oliver.3 Consequently, they provided Fluid-Quip with access to the Dorr-Oliver clamshells located at their plants. Fluid-Quip measured the Dorr-Oliver clamshells and, in August of 1991, marketed its own clamshell with the precise dimensions of Dorr-Oliver's machine. The interior of Fluid-Quip's clamshell is identical to Dorr-Oliver's and utilizes the same starch washing process. Moreover, although there are minor differences in the outer housings of the two manufacturer's clamshells, the district court found, and we agree, that the general external appearance of the two clamshells is practically identical. The Fluid-Quip name, however, is cast into all seven removable sections of the clamshell's outer housing and also appears on a nameplate attached to the housing.

At the time Fluid-Quip entered the market, Dorr-Oliver clamshells sold for about $40,000, while Fluid-Quip charged approximately half that amount. Oftentimes in the industry, customers purchase several clamshells at once, spending hundreds of thousands of dollars. The evidence at trial showed that the twelve companies in the market are very careful when making purchasing decisions regarding clamshells. In the typical clamshell sales process, the supplier initially provides the customer with a preliminary bid. Before deals are closed the supplier usually engages in extensive discussions and negotiations with the customer's upper-level management and engineers, which can last months or even years. In all of its discussions and negotiations, Fluid-Quip clearly informed its potential customers that it was not associated in any way with Dorr-Oliver. Thus, rather than misrepresenting the source of its product, Fluid-Quip truthfully held itself out as an alternative source of clamshells.

Companies in the corn wet milling industry commonly conduct tours of their plants in which people can view clamshells and other equipment involved in the milling process. Often plant managers and engineers participate in reciprocal tours of each other's plants. Additionally, tours have been given at these plants for various international companies in the corn wet milling industry. Representatives of foreign companies also have visited Dorr-Oliver's headquarters, and Dorr-Oliver has led tours of plants for these representatives. No evidence was offered at trial, however, showing that any of these foreign companies have ever purchased a clamshell from either Dorr-Oliver or Fluid-Quip.

Dorr-Oliver filed suit against Fluid-Quip, asserting that, under the Lanham Act and state law,4 it had exclusive rights in the design of the clamshell's outer housing and the use of the word "clamshell" in connection with starch washers. After a six-day bench trial, the district court concluded that Dorr-Oliver had no trademark rights in the word "clamshell." On the other hand, the court determined that Dorr-Oliver did have a protectable trade dress in the outer design of the clamshell, which Fluid-Quip had infringed. The court therefore ruled in favor of Dorr-Oliver on its Lanham Act claim and enjoined Fluid-Quip from selling its clamshell starch washers.5 At the time the injunction was entered, Fluid-Quip had sold eighty-one clamshells, all to customers who owned and operated Dorr-Oliver clamshells. On appeal, Fluid-Quip argues that the evidence presented at trial was insufficient to establish Dorr-Oliver's claim of trade dress infringement.

II.

Section 43(a) of the Lanham Act provides that

15 U.S.C. § 1125(a)(1). Dorr-Oliver seeks protection in the trade dress of its clamshell, which we have defined as "a product's overall image, including its size, shape, color, graphics, packaging, and label." Abbott Labs. v. Mead Johnson & Co., 971 F.2d 6, 20 (7th Cir.1992) (quoting Vaughan Mfg. Co. v. Brikam Int'l, Inc., 814 F.2d 346, 348 n. 2 (7th Cir.1987)). Specifically, Dorr-Oliver claims as its trade dress the configuration of the clamshell itself. See generally Thomas & Betts Corp. v. Panduit Corp., 65 F.3d 654, 657-58 (7th Cir.1995) (discussing implications of patent law on product configuration trade dress cases), cert. denied, ___ U.S. ___, 116 S.Ct. 1044, 134 L.Ed.2d 191 (1996); Kohler Co. v. Moen, Inc., 12 F.3d 632, 642-44 (7th Cir.1993) (granting trademark protection to product configurations does not create an unavoidable conflict with patent law). To prevail on its action for trade dress infringement under § 43(a), a plaintiff must establish (1) that its trade dress has either acquired secondary meaning or is inherently distinctive and (2) that the similarity of the defendant's trade dress to that of the plaintiff causes a likelihood of consumer confusion as to the source or affiliation of the products. See Badger Meter, Inc. v. Grinnell Corp., 13 F.3d 1145, 1151 (7th Cir.1994); Computer Care v. Service Sys. Enters., Inc., 982 F.2d 1063, 1067-68 (7th Cir.1992); Schwinn Bicycle Co. v. Ross Bicycles, Inc., 870 F.2d 1176, 1182 (7th Cir.1989); International Kennel Club of Chicago, Inc. v. Mighty Star, Inc., 846 F.2d 1079, 1087-88 (7th Cir.1988). Even if the plaintiff meets its burden on these two issues, it cannot prevail if the defendant demonstrates that the plaintiff's trade dress is functional. See Badger Meter, 13 F.3d at 1151; Abbott Labs., 971 F.2d at 20. The district court in this case determined that the outer housing of Dorr-Oliver's clamshell was inherently distinctive, had acquired secondary meaning, and was not functional. The court also found that there was a likelihood of consumer confusion based on the similarity of Dorr-Oliver's and Fluid-Quip's clamshells.

Although Fluid-Quip presents a number of compelling arguments for reversal, the only one we need discuss is that the district court erred in finding that Dorr-Oliver had demonstrated a likelihood of consumer confusion.6 The trial court's ultimate determination regarding likelihood of confusion is a finding of fact, which we review for clear error. See, e.g., Forum Corp. of North Am. v. Forum, Ltd., 903 F.2d 434, 438 (7th Cir.1990). Yet "we review the district court's statement of the law de novo for legal error and its conclusions for signs that the court's application of the law was infected with legal error, i.e., an erroneous general principle about the way the test should be applied." Id.; see Schwinn, 870 F.2d at 1187-88.

The requirement of consumer confusion is vital to protecting the basic policies behind federal trademark law. As explained by the Supreme Court,

Park 'N Fly, Inc. v. Dollar Park & Fly, 469 U.S. 189, 198, 105 S.Ct. 658, 663, 83 L.Ed.2d 582 (1985) (internal citation omitted). We have posited that trademark law is designed "to reduce the cost of information to consumers" by facilitating the ability of consumers to identify the producers with which they have had either good or bad experiences. W.T. Rogers Co. v. Keene, 778 F.2d 334, 338 (7th Cir.1985). The law therefore permits a producer to choose an identifier for his product and correspondingly forbids other producers from using confusingly similar identifiers on their products. Id.; see Thomas & Betts Corp. v. Panduit Corp., 65 F.3d 654, 657 (7th Cir.1995). Because a trademark is an identifier rather than a property "right," the use of a competitor's mark that does not cause confusion as to source is permissible. See Libman Co. v. Vining Indus., Inc., 69 F.3d 1360, 1362 (7th Cir.1995), cert. denied, ___ U.S. ___, 116 S.Ct. 1878, 135 L.Ed.2d 173 (1996); August Storck K.G. v. Nabisco, 59 F.3d 616, 618 (7th Cir.1995). Consistent with this economic focus of trademark law, the likelihood of confusion must be determined with reference to the realities of consumer behavior in the relevant market. See Libman, 69 F.3d at 1361; Smith Fiberglass Prods., Inc. v. Ameron, Inc., 7 F.3d 1327, 1329 (7th Cir.1993); Nike, Inc. v. Just Did It Enters., 6 F.3d 1225, 1229 (7th Cir. 1993); James Burrough Ltd. v. Sign of Beefeater, Inc., 540 F.2d 266, 275 (likelihood of confusion must be determined "in the light of what occurs in the marketplace, not the courtroom").

Dorr-Oliver argues that Fluid-Quip attempted to capitalize on Dorr-Oliver's goodwill by confusing consumers into thinking that Fluid-Quip's clamshells were made by, or in some way connected with, Dorr-Oliver. In evaluating the likelihood of consumer confusion for this claim of trade dress infringement, the factors to be considered include: (1) the similarity of the trade dresses; (2) the products to which the trade dresses are attached; (3) the area and manner of concurrent use; (4) the degree of care likely to be exercised by consumers; (5) the strength of the plaintiff's trade dress; (6) actual confusion; and (7) intent of the defendant to pass off its products as those of the plaintiff. See, e.g., Badger Meter, 13 F.3d at 1152. The proper weight given to each of these factors will vary from case to case. Id. The district court mentioned these factors in its opinion, but did not apply them in reaching its conclusion that consumers of clamshells were likely to be confused.

The district court found that there was "no possibility of confusion by the purchaser at or near the point of sale." 894 F.Supp. at 1200. The grounds for this conclusion are apparent, but are certainly useful to discuss. Companies in the corn wet milling industry purchase clamshells directly from the manufacturers, after extensive negotiations and discussions. Fluid-Quip held itself out to this limited market as a competitor of Dorr-Oliver and conspicuously marked its clamshells with its company name. Clearly, any purchasers of clamshells would know from whom they were buying. There is therefore no realistic possibility that anyone would purchase a Fluid-Quip clamshell believing it to be a Dorr-Oliver clamshell.7 See Versa Prods. Co. v. Bifold Co. Mfg. Ltd., 50 F.3d 189, 212 (3d Cir.) (holding that most important factors in evaluating confusion in product configuration cases are marketing and labeling of products), cert. denied, ___ U.S. ___, 116 S.Ct. 54, 133 L.Ed.2d 19 (1995).

Although this analysis at first blush appears to foreclose any finding that consumers are likely to be confused, the district court held that the Lanham Act applies not only to point-of-sale confusion, but also to "post-sale confusion" of potential purchasers. In support of its holding, the district court relied on Lois Sportswear, U.S.A., Inc. v. Levi Strauss & Co., 799 F.2d 867 (2d Cir.1986). In that case Levi Strauss sued the defendants for using a back pocket stitching pattern substantially similar to Levi Strauss' trade-marked stitching pattern. The defendants argued that customers in clothing stores would realize whose product they were buying because the defendants' jeans were clearly labeled at the stores. The court noted that the labeling would not prevent consumers from mistakenly thinking that Levi Strauss was associated with the defendants or had consented to the defendants' use of its trademark. Id. at 872. The court further held that the likelihood of consumer confusion should not be measured solely on the basis of consumers who actually purchase the defendants' jeans believing they were Levi jeans, i.e., solely on point-of-sale confusion:

Id. at 872-73 (emphasis added). Thus, while actual purchasers might have ultimately been informed regarding the source of the defendants' jeans, the court found that the similarity in the stitching patterns would cause prospective purchasers to seek out the defendants' jeans thinking that they were associated with Levi Strauss. Id. at 873. The visibility of the pocket stitching to the consuming public effectively advertised an association of the defendants' jeans with Levi Strauss. After applying its multifactor confusion test and considering the evidence supporting Levi Strauss' claim, the court concluded that there was a likelihood of consumer confusion. Id. at 876. See also Levi Strauss & Co. v. Blue Bell, Inc., 632 F.2d 817, 821-22 (9th Cir.1980) (finding likelihood of consumer confusion on similar facts).

Dorr-Oliver correctly notes that the basic principles expressed in Lois Sportswear can be found within our own precedent. For instance, we have stated that in evaluating the likelihood of consumer confusion, courts should consider "the group of potential purchasers of both products." Forum Corp., 903 F.2d at 442; see International Kennel Club, 846 F.2d at 1087 (likelihood of confusion is based on "probable or actual actions and reactions of prospective purchasers of the goods or services of the parties") (quoting American Int'l Group, Inc. v. London Am. Int'l Corp., Ltd., 664 F.2d 348, 351 (2d Cir.1981)). Thus, the proper inquiry centers on the confusion of consumers in the market for the particular products at issue. See Libman, 69 F.3d at 1364; Smith Fiberglass, 7 F.3d at 1329. In addition, the Lanham Act forbids a competitor from luring potential customers away from a producer by initially passing off its goods as those of the producer's, even if confusion as to the source of the goods is dispelled by the time any sales are consummated. See Forum Corp., 903 F.2d at 442 n. 2; see also Mobil Oil Corp. v. Pegasus Petroleum Corp., 818 F.2d 254, 260 (2d Cir.1987). This "bait and switch" of producers, also known as "initial interest" confusion, will affect the buying decisions of consumers in the market for the goods, effectively allowing the competitor to get its foot in the door by confusing consumers.

The district court found that there was a likelihood of post-sale confusion because various people, including potential foreign purchasers, "would be unable to distinguish Fluid-Quip's clamshells from Dorr-Oliver's when viewing a starch washing line" during tours of industry plants. 894 F.Supp. at 1200. The court reached this conclusion based on the fact that the names cast into the clamshells would not be visible from any distance and photographs of the clamshells showed that some were covered by corn starch dust, making the names difficult to discern. Additionally, the court found it highly relevant that the parts of the outer housings of the two clamshells are interchangeable. Due to this interchangeability, one could expect to find a Dorr-Oliver outer housing on a Fluid-Quip clamshell core. The court reasoned that someone viewing this mixed-manufacturer clamshell in operation might attribute any mechanical problems to Dorr-Oliver, when in fact the problems would be caused by the Fluid-Quip core.

In our judgment, the evidence relied on by Dorr-Oliver and the district court is not sufficient to establish a likelihood of consumer confusion. The proper examination is not whether some people viewing clamshells in industry plants might be confused, but rather whether consumers in the market for clamshells are likely to be confused. Indeed, the Lanham Act is concerned with "customer confusion when choosing to purchase, or not purchase, the items, not public confusion at viewing them from afar." Nike, 6 F.3d at 1229. Although the district court found that plant tours were given to "potential customers" from foreign countries, there is no evidence in the record that anyone other than the twelve domestic companies has ever purchased, or even expressed an intention to purchase, a clamshell. A determination that the market for clamshells includes these foreign visitors would be complete speculation. Indeed, in its brief on appeal, Dorr-Oliver concedes that "[t]he market for starch washers is limited to 12 customers in the corn wet milling industry." It is not possible that any of these companies, all of which own and operate Dorr-Oliver clamshells, will be confused by Fluid-Quip's introduction of a competitive line of clamshells.

Moreover, we would not reach a different result even if we agreed with Dorr-Oliver's position that viewers of operating clamshells are potential clamshell purchasers and should be factored into the likelihood of confusion calculus. Dorr-Oliver's only theory of confusion is that these potential purchasers, upon seeing the two manufacturers' names on the clamshells, will likely believe that Fluid-Quip's clamshells are affiliated with Dorr-Oliver. According to Dorr-Oliver, this will confuse the potential purchasers into having an "initial interest" in Fluid-Quip's product. Yet we fail to see why people viewing two substantially identical clamshells would leap to this conclusion. We believe that, in the context of this industrial machine, the typical consumer will not assume that the two manufacturers are associated in some way. Rather, where product configurations are at issue, consumers are generally more likely to think that a competitor has entered the market with a similar product. See Versa, 50 F.3d at 215. Dorr-Oliver has thus failed to demonstrate "initial interest" confusion. In fact, the district court found that Fluid-Quip marketed its clamshell as an alternative to Dorr-Oliver's. There is no evidence that Fluid-Quip lured or attempted to lure potential customers away from Dorr-Oliver by passing off its clamshells as those of Dorr-Oliver. We must therefore conclude that Dorr-Oliver has not met its burden of establishing a likelihood of consumer confusion and that the district court's determination to the contrary was clearly erroneous.

Ultimately, rather than being grounded in a plausible theory of consumer confusion, the district court's opinion appears to be based on its general notion that "[i]t is inherently unfair for a competitor to enter the market on the back of the originator of a design." 894 F.Supp. at 1200. In the case of product configurations, however, this broad principle runs headlong into the patent laws, which grant limited monopolies for novel and nonobvious products. As a necessary incident to the patent system, "[a]n unpatented article, like an article on which the patent has expired, is in the public domain and may be made and sold by whoever chooses to do so." Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 231, 84 S.Ct. 784, 789, 11 L.Ed.2d 661 (1964). Although we have held that trademark protection can extend to product configurations consonant with the patent laws, a "sensitive application of the principles governing trademark recognition" is necessary to relieve the undeniable tension between the two bodies of law in this area. Kohler, 12 F.3d at 642. In this respect, courts must recognize that "exploiting the goodwill of the article — the attractive features, of whatever nature, that the product holds for consumers — is robust competition; only deceiving consumers, or exploiting the goodwill of another producer, is unfair competition." Duraco Prods. v. Joy Plastic Enters., Ltd., 40 F.3d 1431, 1445 (3d Cir.1994) (internal quotations and citations omitted); see Thomas & Betts, 65 F.3d at 660-61. Thus, while trademark law forbids competitors from copying a product feature that serves as a source identifier, "effective competition and the penumbra of the patent laws require that competitors be able to slavishly copy the design of a successful product." Thomas & Betts, 65 F.3d at 658. One way in which courts have avoided the potential conflict between trademark and patent laws as applied to product configurations is by finding that, in the case of a high-priced, single-purchase product, there is generally no likelihood of confusion when the manufacturer's name is clearly displayed on the product. See, e.g., Versa, 50 F.3d at 213; Merchant & Evans, Inc. v. Roosevelt Bldg. Prods. Co., 963 F.2d 628, 636 (3d Cir.1992); Litton Sys. v. Whirlpool Corp., 728 F.2d 1423, 1446-47 (Fed.Cir.1984); Bose Corp. v. Linear Design Labs, Inc., 467 F.2d 304, 309-10 (2d Cir.1972).

In the current case Dorr-Oliver reaped the rewards of its patents on the clamshell for seventeen years, after which time the product passed into the public domain. Fluid-Quip then entered the clamshell market with a product virtually identical in appearance and represented itself to potential customers as a competitor of Dorr-Oliver. This is certainly competition, but the similarity of the two clamshells, standing alone, does not make it unfair. Indeed, the "mere inability of the public to tell two identical articles apart is not enough to support an injunction against copying or an award of damages for copying that which the federal patent laws permit to be copied." Stiffel, 376 U.S. at 232, 84 S.Ct. at 789. The district court's analysis simply neglected to account for the delicate interplay of the patent and trademark laws in the context of product configurations.8

III.

We have held that "a finding of likely confusion can no more be based on pure conjecture or a fetching narrative alone than any other finding on an issue on which the proponent bears the burden of proof." See Libman, 69 F.3d at 1363. Dorr-Oliver has not presented a plausible theory of consumer confusion, let alone one supported by evidence. Hence, Dorr-Oliver has failed to establish its claim of trade dress infringement under § 43(a) of the Lanham Act. The district court relied on the likelihood of confusion finding in ruling for Dorr-Oliver on its state law unfair competition claims. Dorr-Oliver does not argue that its state law claims should be treated differently than its claim for trade dress infringement under the Lanham Act. We therefore REVERSE the judgment of the district court in favor of Dorr-Oliver on the Lanham Act and state law claims and REMAND the case for further proceedings consistent with this opinion.