Krueger Intern., Inc. v. Nightingale Inc.

Full title: KRUEGER INTERNATIONAL, INC., Plaintiff, v. NIGHTINGALE INC., Defendant

Court: United States District Court, S.D. New York

Date published: Jan 31, 1996

915 F. Supp. 595 (S.D.N.Y. 1996)

Opinion

No. 95 Civ. 10703 (SS).

January 31, 1996.

Kenyon Kenyon, New York City, (Allen J. Baden, of counsel), for Plaintiff.

Darby Darby P.C., New York City, (Andrew Baum, of counsel), for Defendant.

OPINION AND ORDER

SOTOMAYOR, District Judge.

Plaintiff, a manufacturer of metal-frame stacking chairs, alleges that defendant, a competitor, has "slavishly copied" its chair design and thereby infringed its trade dress under § 43(a) of the Lanham Act, 15 U.S.C. § 1125(a).¹ The plaintiff, Krueger International, Inc. ("KI"), moves for a preliminary injunction against the defendant, Nightingale Inc. ("NI"), on the grounds that the KI design is distinctive and that NI's design is confusingly similar. KI seeks to enjoin NI from advertising, manufacturing or selling copies of the KI chair. For the reasons discussed below, KI's motion is DENIED.

1 Plaintiff does not cite the Lanham Act or any particular section of that Act in its Notice of Motion. Plaintiff does mention the Act in a passing reference in its supporting papers. (Mem. in Support of Motion at 22). I assume for the purposes of this motion that plaintiff means to invoke 15 U.S.C. § 1125(a), which has been interpreted as providing protection to unregistered trade dress designs. See LeSportsac, Inc. v. K Mart Corp., 754 F.2d 71, 75 (2d Cir. 1985).

BACKGROUND

The parties to this action are established manufacturers of office and industrial furniture. KI, founded in 1941, is headquartered in Green Bay, Wisconsin. NI is a Canadian corporation established in 1928. Both companies produce numerous lines of chairs for commercial use. At issue here is KI's Matrix² chair, which competes in the "high density" stacking chair market. High density chairs are strong, lightweight chairs, capable of being stacked up to 40 units high. The stacking feature of the chairs makes them easy to transport and efficient to store. The chairs are sold in large volume, often through a competitive bidding process, to convention centers, school systems and other institutions that must accommodate crowds. There is no direct retail market for the chairs, although some of these chairs are sold to furniture dealers.

The Matrix chair was introduced in 1976, received a design patent in 1978, and quickly became KI's single most successful product.³ At least 10 other designs currently compete in this market.⁴ Although the Matrix design patent expired in 1991, no competitor has attempted to produce an exact copy of the Matrix chair until the NI chair.

3 KI asserts that Matrix sales (at wholesale price) have totaled more than $150 million, including almost $50 million since 1991. (Pl.'s Notice of Motion at 5.)

4 In their motion papers, the parties have introduced photographs of other competing products.

This Court held oral argument on January 17, 1996, at which the parties produced specimens of their own chairs as well as those of two competitors. The defendant's chair, called the Beetle 300, is a virtual carbon copy of the Matrix chair. The other competing designs are visually distinguishable. Although they all have a thin metal frame made of chrome-covered steel, and a seat and back made of molded plastic, the Matrix and Beetle chairs share three other features which KI claims as trade dress. One feature is a pair of Z-shaped metal rods that zig-zag in and out, connecting the seat and the back. (Competitors use straight connecting rods between the seat and back.) The second feature is the shape of the seat, which has an unusually deep "dip," and the third is the shape of the back, which suggests a trapezoid.

Defendant NI frankly acknowledges that it copied the Matrix chair down to the last detail, and states that it did so because KI's design patent had expired and because NI needed to clone the Matrix chair in order to meet government specifications in a competitive bidding process. The bids had been solicited by a government-owned corporation, Unicor, which buys chair components from private manufacturers and sends them to federal prisons, where inmates assemble them. Unicor then sells the completed chairs to federal government agencies. Until July 1994, KI was the fortunate holder of the Unicor contract. When the contract expired, Unicor began advertising for new bids. Included in the bid specifications was a technical drawing of the Matrix chair. Because much of the subsequent story turns on what KI knew about defendant's response to those specifications, and when, it is appropriate to set out the facts of the bid process in some detail.

At a pre-bid conference organized by Unicor in December, 1994, Gregory Wallis, a Unicor executive, opened the meeting by addressing Unicor's general concerns, including the following:

One issue that we typically are concerned about is if there is any kind of patent or proprietary rights for the products that we are buying. We have had some history with that and we want to be made aware of that up front if there is anything that we would not be allowed to either produce ourselves or purchase elsewhere and combine with maybe your component parts, something of that nature. If there is any kind of restrictive rights under the products that we will be buying, we need to know that up front.

(Tr. of Unicor Pre-Bid Conf. at 13-14 (emphasis added).) Later in the same meeting, another Unicor spokesman, Michael Campanale, answered five written questions from prospective bidders. Two of the five questions concerned the Matrix chair and its prominent place in Unicor's specifications.

Q: The RFP [Request for Proposal] indicates that all offerors will be evaluated per the minimum specification requirements. Yet many of the [technical requirements] are either proprietary or highly restrictive. Are we to propose what we have that we believe is best for Unicor or are we to force our products to meet Krueger Manufacturing methods and specifications? To do so will cause higher costs for us and a significant competitive advantage to Krueger.

A: All of the indicated specifications and requirements are not intended to be restrictive. This is exactly why we developed it as an RFP, Request for Proposal, and not simply as a bid package. All samples will be considered that fall within a competitive range of pricing and technical factors. . . .

Q: Section C [of the Request] is obviously a set of specifications for the Krueger Matrix . . . chair. Are these considered hard specifications or can a product which does not meet these exact specifications, but is an equal or even superior product, be proposed and awarded?

. . . . .

A: [Y]es. All proposals which indicate equal or even superior specifications will be considered.

( Id. at 20-21, 23.) Campanale explained that the specifications were meant to ensure quality and to maintain "a general consistency with the hundreds of thousands of high density stacking chairs" (i.e., Matrix chairs) that Unicor had already sold to federal agencies. "There is also a need to consider the customer's concern and possible preference to be able to have chairs which meet, reasonably match in the areas of color, stackability, style, quality and performance." ( Id. at 24-25.) Two employees of KI attended this conference.

In March 1995, Unicor awarded the contract to NI, having ranked KI's bid third, below competitor CPSI. The contract is worth at least $7.7 million over five years.

Unicor immediately asked NI to provide a prototype of the Beetle chair to allow Unicor to verify that the Beetle and Matrix chairs would "stack" interchangeably. (Campanale Decl. at 2.) Within a few days, a prototype arrived at the Unicor office. The chair was a patched-together affair: NI had made the metal frame, but had cannibalized the seat and back sections from a Matrix chair. As NI states in its motion papers, "[t]he prototype included Nightingale's own frame, but because we had not yet tooled the back and seat, we inserted a Matrix back and seat into the prototype, for purposes of the prototype only." (Breen Decl. at 5.) Significantly, the NI frame featured two Z-shaped rods connecting the seat and back. The Matrix components fit on this frame perfectly.

A few days later, KI Vice President Daniel Schiltz visited Unicor and saw the Beetle prototype. (Schiltz Aff. at 1.) Schiltz was sufficiently alarmed to protest to Campanale, a Unicor executive, that "it appeared to be a KI chair." (Campanale Decl. at 2.) Campanale said that the chair was only a prototype and that NI would soon be making its own parts for the seat and back. ( Id.)

On March 30, 1995, KI filed a formal protest with the Government Accounting Office (the "GAO"), challenging the contract award to NI. The essence of KI's protest was that Unicor had improperly favored NI in the bidding process by waiving certain requirements for NI but not for other bidders, and had given more weight to price considerations than the specifications had indicated. KI also protested the fact that Nightingale had submitted a prototype using KI parts.

KI's protest effected an immediate stay of the contract and Unicor directed NI to stop production until further notice. In the meantime, NI began preparing photographs of the Beetle chair for promotional purposes. In May 1995, as part of the bid protest process, NI sent an updated copy of its entire chair product line to the GAO and to KI's attorney. (Kunzi Decl. at 1.) Included was a photograph of the Beetle chair. Although a KI lawyer knew what the Beetle chair looked like as of May 1995, the lawyer maintains that this photograph was subject to a GAO protective order, which barred attorneys from disclosing competitive information to their clients during the bid protest process. NI's lawyer disagrees, contending that a publicity brochure does not constitute competitive information subject to the protective order.

Wholly apart from the bid protest, however, NI soon began sending new product catalogues to its representatives and dealers. (Breen Decl. at 6.) These catalogues also contained photographs of the Beetle chair. In June, a shipment of catalogues arrived at Unicor, where Campanale placed them on a table. Sometime thereafter, Schiltz, the same KI executive who had seen the Beetle prototype and had expressed concern, visited Campanale. This time Schiltz looked through the catalogue, "noticed the photographs of the Beetle chair," and took a copy with him. (Campanale Decl. at 2.)

In July 1995, although the bid protest was still pending, Unicor instructed NI to begin immediate performance on the contract "in the best interests of the Government." In September, Unicor notified both the GAO and KI's attorney that NI had begun production. In October, the GAO dismissed KI's bid protest because Unicor had agreed to reopen bidding on technical considerations. Between September and December, 1995, Nightingale shipped enough chair parts to Unicor to produce 10,000 Beetle chairs. Its contract, however, is still subject to termination if its product samples fail Unicor testing. (Ralston Decl. at 3.)

In November 1995, Nightingale displayed the Beetle chair at trade shows in New York City and Toronto. It was at these trade shows, the first on November 2 and the second on November 17, KI asserts, that it first saw the Beetle chair. Not until December 7, 1995, however, did an investigator for KI attempt to find a Beetle chair and not until about December 11, 1995, was KI's Canadian counsel successful in securing a chair. (Jaraczewski Aff. at 6; Jameson Aff. at 1-2.) KI filed a Complaint in this action on December 19, 1995, and requested a preliminary injunction two days later.

DISCUSSION

I. Merits of the Lanham Act Claim

A. Inherent Distinctiveness.

This case presents one of the most difficult analytical issues in all of trade dress law: how to determine whether a product design is "inherently distinctive." The issue is particularly pressing and unresolved in this circuit, whose rule for trade dress protection was upended three and a half years ago by the Supreme Court in Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 112 S.Ct. 2753, 120 L.Ed.2d 615 (1992). Prior to Two Pesos, a plaintiff seeking trade dress protection in this circuit had to prove that the dress had acquired secondary meaning. Two Pesos, which unified the standard for trademark and trade dress law, held that protection for trade dress requires either inherent distinctiveness or secondary meaning, but not both. Two Pesos, 505 U.S. at 774, 112 S.Ct. at 2760.

The general rules of trade dress protection had been long established. Section 43(a) of the Lanham Act protected the "trade dress" of a product, which is defined as the product's "total image," including "features such as size, shape, color or color combinations, texture, [or] graphics." LeSportsac, Inc. v. K Mart Corp., 754 F.2d 71, 75 (2d Cir. 1985) (citations omitted). It was equally well settled that trade dress includes both product packaging and product design, and that trade dress is protected only when the infringer's product threatens to confuse customers about the source of the goods. Id. Two Pesos clarified the standard for trade dress protection as follows: a plaintiff must show (1) that its trade dress is distinctive, either because it is inherently so or because it has acquired secondary meaning, and (2) that there is a likelihood of confusion between its product and the defendant's product. Tough Traveler, Ltd. v. Outbound Products, 60 F.3d 964, 967 (2d Cir. 1995) (citing Two Pesos, 505 U.S. at 770, 112 S.Ct. at 2758). A defendant may avoid liability, however, by showing that the trade dress is functional. Two Pesos at 770, 112 S.Ct. at 2758.

The classic test for determining the distinctiveness of a trademark was outlined in Abercrombie Fitch Co. v. Hunting World, Inc., 537 F.2d 4, 9 (2d Cir. 1976), and the test has since been extended to trade dress as well. See Jeffrey Milstein, Inc. v. Greger, Lawlor, Roth, Inc., 58 F.3d 27, 31 (2d Cir. 1995); Paddington Corp. v. Attiki Importers Distribs., Inc., 996 F.2d 577, 583-84 (2d Cir. 1993); accord, Two Pesos, 505 U.S. at 773-74, 112 S.Ct. at 2760. Under the Abercrombie test, marks are classified as either (1) generic, (2) descriptive, (3) suggestive, or (4) arbitrary or fanciful. Abercrombie, 537 F.2d at 9. Suggestive, arbitrary, and fanciful marks and dresses are always considered inherently distinctive, in that their "intrinsic nature serves to identify a particular source of a product" or they are " capable of identifying a particular source of the product," whether or not the trade dress has acquired secondary meaning or a wide public association with the source. Two Pesos at 767, 771-73, 112 S.Ct. at 2757, 2759 (emphasis added). Descriptive marks are protected only if they have acquired secondary meaning. Generic marks receive no protection. Id. at at 767-69, 112 S.Ct. at 2757-58.

But as many courts have observed in varying degrees of frustration, the Abercrombie classifications do not translate easily to the trade dress context. The problem is not severe when the trade dress involves product packages and labels, which, like trademarks, at least have the advantage of using words and symbols independent of the product to convey information to consumers. See, e.g., Paddington, 996 F.2d at 583-84 (holding No. 12 Ouzo label and bottle to be inherently distinctive). The problem is often daunting, however, when it comes to product designs. Does the shape of a chair seat "suggest" a chair seat? Does it "describe" a chair seat? Or is it just a chair seat? No matter how beautifully designed, an industrial product is what it is. This conundrum recently led the Second Circuit in Knitwaves, Inc. v. Lollytogs Ltd., 71 F.3d 996 (2d Cir. 1995), to abandon the Abercrombie test for product design and announce a new test. I believe, however, that this new test is not particularly helpful because it does not clearly address the standards in this area as set by the Supreme Court.

In Knitwaves, this circuit announced a departure from its earlier approach to trade dress law. Knitwaves involved the design of children's sweaters, particularly the use of "leaf" and "squirrel" designs placed on the sweaters. The court stated (1) that the Abercrombie classifications "make little sense when applied to product features" and were therefore "inapplicable" to product designs, and (2) that henceforth the new test for inherent distinctiveness is whether the manufacturer "used" or "intended to use" the design "to identify the source and distinguish his or her goods." Knitwaves at 1006-09 (citations omitted). In adopting this approach, the court cited approvingly to the Third Circuit decision in Duraco Prods., Inc. v. Joy Plastic Enterprises, Ltd., 40 F.3d 1431, 1440-41 (3d Cir. 1994). The Duraco approach, however, leaves much to be desired. Cf. Tough Traveler, 60 F.3d at 970 n. 1 (Jacobs, J., concurring) ("Under no circumstances do I believe this appeal compels us to decide whether . . . we should adopt the method of analysis" in Duraco). The Eighth Circuit instead has recently made a forceful and persuasive argument that the Supreme Court has not authorized us to abandon Abercrombie, no matter how much difficulty it causes. Stuart Hall Co. v. Ampad Corp., 51 F.3d 780, 788 (8th Cir. 1995). As the Eighth Circuit reasoned, Two Pesos was clearly a case of product design and it expressly approved the application of the Abercrombie classifications to the design of a Mexican restaurant chain. Two Pesos, 505 U.S. at 773-74, 112 S.Ct. at 2760. Moreover, the entire thrust of Two Pesos was to unify the standards for trademark and trade dress, not to balkanize this complex field into yet more subcategories. I agree with the Eighth Circuit's conclusion that the Supreme Court envisions trade dress as a "single concept" with trademark law requiring a single test for inherent distinctiveness. Stuart Hall, 51 F.3d at 788.

Further, I believe that Knitwaves's new test for inherent distinctiveness confuses the analytical requirements for inherent distinctiveness with those of secondary meaning. Inherent distinctiveness cannot hinge on how a producer intends to promote a design. If such were the case, the evidentiary requirements for inherent distinctiveness would be almost identical to those for secondary meaning, and there would be no point in having two categories. A producer could only prove "intent" by producing evidence of how he or she has advertised (or positioned) the product, and by producing consumer surveys showing how well the advertising worked. This is precisely the kind of evidence required for a showing of secondary meaning. See Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 851 n. 11, 102 S.Ct. 2182, 2187 n. 11, 72 L.Ed.2d 606 (1982) ("To establish secondary meaning, a manufacturer must show that, in the minds of the public, the primary significance of a product feature . . . is to identify the source of the product rather than the product itself.") (emphasis added). An inherent quality, by contrast, is one which "inheres" in a product, regardless of what the producer intends, particularly if the product is too new to have acquired secondary meaning. Two Pesos created no such "intent" requirement, stating only that an inherently distinctive design is one that is " capable of identifying a particular source of the product." Two Pesos, 505 U.S. at 771, 112 S.Ct. at 2759 (emphasis added). Furthermore, the Court explicitly cautioned that "[e]ngrafting onto § 43(a) a requirement of secondary meaning for inherently distinctive trade dress . . . would undermine the purposes of the Lanham Act." Id. at 774, 112 S.Ct. at 2760.

In fashioning its "intent" requirement, the Knitwaves court relied on dicta from the Supreme Court's opinion in Qualitex Co. v. Jacobson Products Co., ___ U.S. ___, ___, 115 S.Ct. 1300, 1303, 131 L.Ed.2d 248 (1995), but Qualitex was a trademark case that hinged entirely on secondary meaning. In Qualitex, the issue was whether a color could serve as a registered trademark. Justice Breyer did quote directly from the Lanham Act, 15 U.S.C. § 1127, which does indeed define a trademark as any "word, name, symbol, or device" that a person "use[s]" or "inten[ds] to use . . . to identify and distinguish his or her goods." Id. But his inquiry did not stop there. The Court then asked whether a color could be inherently distinctive. Upon determining that it could not, the Court proceeded to observe that "over time, customers may come to treat a particular color . . . as signifying a brand." Qualitex, ___ U.S. at ___, 115 S.Ct. at 1303. This is secondary meaning. The Court concluded that the green-gold color at issue in Qualitex, although not inherently distinctive, had acquired secondary meaning and therefore could be registered as a trademark. Id. Hence, a close reading of Qualitex does not support the notion that the "use or intended use" language of the Lanham Act provides a definition of inherent distinctiveness.

Knitwaves, however, does provide some guidance for a workable approach to inherent distinctiveness, particularly when read in conjunction with efforts by other courts. The Fifth and Eighth Circuits offer particularly useful instruction. The Fifth Circuit's Chevron test, first developed for product packaging, states that "[i]f the features of the trade dress sought to be protected are arbitrary and serve no function either to describe the product or assist in its effective packaging," the dress is inherently distinctive. Chevron Chemical Co. v. Voluntary Purchasing Groups, Inc., 659 F.2d 695, 702 (5th Cir. 1981) (packaging of lawn and garden products), cert. denied, 457 U.S. 1126, 102 S.Ct. 2947, 73 L.Ed.2d 1342 (1982). This is essentially a test of functionality. The Supreme Court explicitly approved the application of the Chevron test to product designs in its decision in Two Pesos. 505 U.S. at 773-74, 112 S.Ct. at 2760.

The Eighth Circuit also emphasizes the issue of functionality, asking:

whether, and how much, the trade dress is dictated by the nature of the product. . . . If the specific design of the trade dress is only tenuously connected with the nature of the product, then it is inherently distinctive. . . . If the design of the trade dress is dictated by the nature of the product, then secondary meaning must be proven.

Stuart Hall, 51 F.3d at 786.

A third test based on the Abercrombie classifications was developed 19 years ago by the Court of Customs and Patent Appeals:

In determining whether a design is arbitrary or distinctive this court has looked to whether it was a `common' basic shape or design, whether it was unique or unusual in a particular field, [or] whether it was a mere refinement of a commonly-adopted and well-known form of ornamentation for a particular class of goods view by the public as a dress or ornamentation for the goods.

Seabrook Foods, Inc. v. Bar-Well Foods Ltd., 568 F.2d 1342, 1344 (C.C.P.A. 1977). The Seabrook test is the most useful of the three tests, in my view, because it clarifies the importance of market context. The Seabrook test asks "whether the design, shape or combination of elements is so unique, unusual or unexpected in this market that one can assume without proof that it will automatically be perceived by customers as an indici[um] of origin." 1 McCarthy, Trademarks and Unfair Competition § 8.02[4] (emphasis added); see also AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531 (11th Cir. 1986) (applying Seabrook test to product design), cert. denied, 481 U.S. 1041, 107 S.Ct. 1983, 95 L.Ed.2d 822 (1987). In other words, Seabrook highlights the notion that a design cannot be considered in a vacuum, which courts have known all along. See, e.g., Thompson Medical Co. v. Pfizer Inc., 753 F.2d 208, 213 (2d Cir. 1985) ("the determination whether a mark is descriptive or suggestive cannot be made in a vacuum"). Any test of inherent distinctiveness must ask, "Inherently distinctive as compared to what?" The various tests of functionality, discussed in greater detail infra, serve much the same purpose; they ask whether a given design is functionally necessary in a given market (and therefore common) or whether it is more akin to a hood ornament. See, e.g., W.T. Rogers Co. v. Keene, 778 F.2d 334, 339 (7th Cir. 1985) (Posner, J.). The Fifth Circuit implicitly asked similar questions in Two Pesos when it considered whether, within the universe of Mexican restaurant chains, the plaintiff's particular restaurant decor was common or uncommon. Taco Cabana Int'l, Inc. v. Two Pesos, Inc., 932 F.2d 1113, 1118-19 (5th Cir. 1991), aff'd, 505 U.S. 763, 112 S.Ct. 2753, 120 L.Ed.2d 615 (1992).

The Seabrook approach is fully consistent with the case law in the Second Circuit. Before Two Pesos, the question of market context generally arose in light of the so-called functionality defense. E.g., Wallace Int'l Silversmiths, Inc. v. Godinger Silver Art Co., 916 F.2d 76, 79-81 (2d Cir. 1990) (no protection for widely-used elements of baroque design in silverware), cert. denied, 499 U.S. 976, 111 S.Ct. 1622, 113 L.Ed.2d 720 (1991); Stormy Clime Ltd. v. ProGroup, Inc., 809 F.2d 971, 977 (2d Cir. 1987) (no protection for functional elements of rain jackets). In the wake of Two Pesos, this circuit has continued to apply a Seabrook-type analysis to the puzzle of inherent distinctiveness. See, e.g., Mana Products, Inc. v. Columbia Cosmetics Mfg., Inc., 65 F.3d 1063, 1069-70 (2d Cir. 1995) (examining "the custom in a particular industry" for packaging cosmetics in black containers); Paddington, 996 F.2d at 583 (citing examples of industry customs: "packaging lime-flavored soda in green twelve-ounce cans is so common in the soft drink industry that such packaging probably is not inherently distinctive"). In Knitwaves, the court examined the clothing market and concluded that when people buy clothing, they do not tend to use overall design as an indicator of source. Knitwaves at 1008-09. This approach followed naturally from Stormy Clime, the case involving rain jackets, in which the court declared that labels, more than design, provide a "traditional means of identifying the manufacturer of clothing." Stormy Clime, 809 F.2d at 977. Knitwaves also quoted approvingly from the Restatement (Third) of Unfair Competition, which asks whether "because of the nature of the [design] and the context in which it is used, prospective purchasers are likely to perceive it" as a source-identifier. Knitwaves at 1008-09 (citing Restatement (Third) of Unfair Competition § 13(a) (1995) (emphasis added)). When considered in this light, Knitwaves's emphasis on the producer's "intent" may be analogous to asking, "What is the custom in this industry?" Or: "How does this industry, including this claimant, use designs?" See also Mulberry Thai Silks, Inc. v. K K Neckwear, Inc., 897 F. Supp. 789, 796-97 (S.D.N.Y. 1995) (in men's necktie industry "[t]he use of geometric design elements, bright colors and the like is common"); Casa Editrice Bonechi, S.R.L. v. Irving Weisdorf Co., Ltd., 1995 WL 528001 at *6 (S.D.N.Y. 1995) (applying Seabrook test to souvenir guidebooks); Jaret Int'l, Inc. v. Promotion In Motion, Inc., 826 F. Supp. 69, 75 (E.D.N Y 1993) (applying Seabrook test to sour candies); EFS Marketing, Inc. v. Russ Berrie Co., 836 F. Supp. 128 (S.D.N.Y. 1993) (no trade dress protection for "troll" doll with pointed ears, pug nose, pot belly, wild hair and outstretched arms, because such elements are common in doll industry).

Before applying these principles to the instant case, it is appropriate to address some common misconceptions about the interplay of patent law and trade dress law. The plaintiff in this case contends that because the Matrix chair has received a design patent, this patent constitutes proof of inherent distinctiveness. The defendant argues just the opposite: that KI is unlawfully seeking to extend its expired patent monopoly, and that patented designs can never claim the protection of trade dress law. Both arguments are incorrect. A design patent is analytically distinct from a protectable trade dress, and industrial products may qualify for both kinds of protection without violating the policy goals of either patent or trade dress law. See, e.g., Hubbell Inc. v. Pass Seymour, Inc., 883 F. Supp. 955, 959-60 (S.D.N.Y. 1995) (Sweet, J.); Rogers, 778 F.2d at 337; 1 McCarthy, Trademarks and Unfair Competition § 6.01[5] (3d ed. 1995). The purpose of patent law is to encourage innovation by excluding competition (i.e., copying) for a limited period of time. 1 McCarthy § 6.01[5]. The purpose of trademark law is to prevent confusion as to the source of goods. 1 Id. § 6.03[5] (citations omitted). As the Court of Customs and Patent Appeals has observed:

[T]rademark rights . . . do not "extend" the patent monopoly. They exist independently of it, under different law and for different reasons. The termination of either has no legal effect on the continuance of the other. When the patent monopoly ends, it ends. The trademark rights do not extend it.

In re Mogen David Wine Corp., 328 F.2d 925, 930 (C.C.P.A. 1964) [Mogen David I].

It is also important to note that the requirements for a design patent are distinct from those for a trade dress. A design patent (unlike a utility patent) is granted for a product's non-functional, ornamental features. It is appropriate only for industrial design, and one of its purposes is to promote progress in the "decorative arts." 1 McCarthy § 6.03[5]. To qualify for a design patent, the features must be new, non-obvious and ornamental. 1 Id. They need not be inherently distinctive. As the Supreme Court has stated, "To qualify for protection, a design must present an aesthetically pleasing appearance that is not dictated by function alone, and must satisfy the other criteria of patentability." Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 148, 109 S.Ct. 971, 976, 103 L.Ed.2d 118 (1989).

When a design patent expires, the design becomes copyable. It may not, however, be copied in such a way that customers are deceived about what they are buying. "Federal trademark laws, independent in origin from design patent law, have the dual purpose of protecting both the trademark owner and the public from confusion, mistake and deception." In re Mogen David Wine Corp., 372 F.2d 539, 543 (C.C.P.A. 1967) (Smith, J., concurring) [Mogen David II]. Trademark rights last for an indefinite period, depending on the strength and duration of the public's recognition that the mark identifies the producer's goods. Id. Novelty is irrelevant for trademark or trade dress protection; the design must meet the more rigorous test of (1) inherent distinctiveness or secondary meaning, and (2) likelihood of confusion. Many patented (or formerly patented) designs cannot meet this standard. If the producer has done nothing to market the product, the producer will be unable to show either secondary meaning or likelihood of confusion. If, by contrast, the design has so dominated the market that it virtually defines it, the design may be generic or functional and thus ineligible for protection. See, e.g., Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 59 S.Ct. 109, 83 L.Ed. 73 (1938) (pillow shape of shredded wheat biscuit, being primarily functional, fell into public domain). If the market has evolved into one in which knock-offs or close copies are common and consumers are sophisticated, as in the clothing and handbag industries, the likelihood of confusion will likely be low. E.g., Coach Leatherware Co. v. AnnTaylor, Inc., 933 F.2d 162, 168-70 (2d Cir. 1991) (buyers of Coach handbags know to look for Coach registration number and Coach name). "Trademark law is replete with instances where the public has both `created' and `destroyed' the trademark rights which are subject to registration." Mogen David II, 372 F.2d at 543. A design patent, therefore, can neither guarantee nor preclude a finding of protectable trade dress.

The existence of a design patent, however, is relevant to the functionality defense. Because a design patent is granted only for non-functional designs, it can serve as evidence that a plaintiff's trade dress is not functional. See, e.g., In re Morton-Norwich Prods., Inc., 671 F.2d 1332, 1342 n. 3 (C.C.P.A. 1982) (design patent presumptively indicates that design is not de jure functional); In re R.M. Smith, Inc., 734 F.2d 1482, 1485 (Fed. Cir. 1984) (existence of design patent "may be some evidence of non-functionality"). This is not tantamount to saying, however, that a design patent always serves as a trademark. See, e.g., R.M. Smith, 734 F.2d at 1485 (existence of design patent "does not, without more, bestow upon said device the aura of distinctiveness or recognition as a trademark") (citation omitted); In re Vico Prods. Mfg. Co., 229 U.S.P.Q. 364, 370 (T.T.A.B. 1985) (quoting R.M. Smith). At best, a design patent can only help rebut the functionality defense; it cannot do the whole job of proving inherent distinctiveness. The plaintiff still needs to show that, in its market at the time of the alleged infringement, the design is both ornamental and deserving of trade dress protection.

The functionality doctrine was articulated by the Supreme Court in Inwood Laboratories v. Ives Laboratories, 456 U.S. 844, 850 n. 10, 102 S.Ct. 2182, 2187 n. 10, 72 L.Ed.2d 606 (1982), which instructed that "[i]n general terms, a product feature is functional if it is essential to the use or purpose of the article or if it affects the cost or quality of the article." The Second Circuit elaborated on this definition by stating that a design feature is essential "only if the feature is dictated by the functions to be performed; a feature that merely accommodates a useful function is not enough." Warner Bros., Inc. v. Gay Toys, Inc., 724 F.2d 327, 331 (2d Cir. 1983). The doctrine serves to promote competition by "protect[ing] advances in functional design from being monopolized" and "encourag[ing] . . . the broadest dissemination of useful design features." Id. Put in layman's terms, one might ask: "Is the article in this particular shape for utilitarian reasons," i.e., because it works better that way? 1 McCarthy § 7.26[3][b]. In Stormy Clime, this circuit suggested three factors to be considered:

the degree of functionality of the similar features, the degree of similarity between non-functional (ornamental) features of the competing products, and the feasibility of alternative arrangements of functional features that would not impair the utility of the product. These factors should be considered along a continuum. On one end, unique arrangements of purely functional features constitute a functional design. On the other end, distinctive and arbitrary arrangements of predominantly ornamental features that do not hinder potential competitors from entering the same market with differently dressed versions of the product are non-functional and hence eligible for trade dress protection. In between, the case for protection weakens the more clearly the arrangement of allegedly distinctive features serves the purpose of the product . . ., especially where the competitor copying such features has taken some significant steps to differentiate its product.

Stormy Clime, 809 F.2d at 977.

The problem, of course, is that a design feature may be both functional and pleasing. (Indeed, many industrial designs combine these two elements.) This has led the courts into a hopeless tangle about the difference between aesthetics and functionality. Some courts developed a doctrine of "aesthetic functionality" that was both unnecessary and illogical. If a pleasing design helped to sell a product, so the theory went, its aesthetic appeal became an "essential" competitive quality that could not in fairness be denied to others. Thus the doctrine of "aesthetic functionality" denied trade dress protection to design features whose only sin was to delight the senses. In 1990, however, the Second Circuit substantially redefined this doctrine and injected a welcome dose of sanity to the discussion. Wallace, 916 F.2d at 80-81 (Winter, J.). In Wallace, Judge Winter built on the principles of Stormy Clime and its focus on preserving healthy competition. As Judge Newman had stated in Stormy Clime, the availability of alternative designs should guide the inquiry. Stormy Clime, 809 F.2d at 978. Applying this reasoning to the silverware market, Judge Winter found that baroque designs were so common in that industry that they could not be given trade dress protection without "excluding competitors from a substantial market." Wallace, 916 F.2d at 81. In other words, a pleasing design per se is no bar to protection. The plaintiff in Wallace had merely tried to monopolize too broad an aesthetic category, seeking protection "not for a precise expression of a decorative style, but for basic elements of a style that is part of the public domain." Id.

The approach in Knitwaves, however, seems to be a step backward from Wallace because it revives a false dichotomy between aesthetics and source identification. Knitwaves seems to instruct that a design can serve only one primary purpose: either aesthetic or source-identifying, but not both. Knitwaves at 1008-09 (no protection because plaintiff's "objective in the two sweater designs was primarily aesthetic," and therefore "the designs were not primarily intended as source identification"). This approach is neither helpful nor logical. In the instant case, it would force me to determine whether the plaintiff chose its chair design because it was attractive or because it was meant to serve as a trademark (i.e., a source identifier). The likely answer is, both. Moreover, this plaintiff never had to choose between the two objectives because the design was patented and could not be copied. In more general terms, every producer hopes to create a design that is more pleasing than not and that that design will be associated with it. Every producer hopes that it and its trade dress will be remembered because its design is pleasing, not because it is ugly. See Rogers, 778 F.2d at 340 ("Rare is the manufacturer who will not try to choose a pleasing name, symbol, or design feature as his trademark.") Thus I see no need to force the courts to choose between ideas that are not mutually exclusive. The more useful approach, I believe, is the market foreclosure test articulated in Wallace. See also 1 McCarthy § 7.26[3][e] ("the availability of alternative designs is a very important, if not the most important, evidentiary key to unlocking the mystery of the functionality puzzle").

Applying these principles to the facts before me, I conclude that plaintiff KI has a protectable trade dress in the overall look of its Matrix chair. Although each of the individual design elements serves both functional and aesthetic purposes, it is the overall look that I must consider. Milstein, 58 F.3d at 32 ("Although each element of a trade dress individually might not be inherently distinctive, it is the combination of elements that should be the focus of the distinctiveness inquiry."). I base my following conclusions on the photographs of competing chairs submitted by the parties and the accompanying affidavits describing the high density stacking chair market. On this evidence, I find that the seat, the back and the Z-shaped connecting rods of the KI chair give the Matrix chair a distinctive overall look.⁵ Applying Seabrook, I find that this particular look is unique among high density stacking chairs and that manufacturers of these chairs generally seek unique designs as an important source identifier. Unlike the clothing industry, in which labels are prominent because designs change frequently, the chair industry does not rely heavily on labels to communicate with customers and its designs remain stable over time. Although the chair manufacturers' names appear prominently on shipping containers and hang tags, customers generally throw these away. The sole permanent label is molded into the plastic underneath the seat, where a sophisticated buyer would find it but a less sophisticated one might not. No consumer would see the label without looking for it.

5 Plaintiff has submitted the Declaration of Gregory. B. Wallis, a Contracting Officer at Unicor, to support its argument that the Unicor specifications were minimum requirements and did not necessitate copying KI's chair. Nevertheless, ¶ 8 of the Wallis Declaration appears to raise a question about the distinctiveness of the Matrix chair: "Krueger's products were determined to be generally representative of the market and was (sic) used as a baseline reference. To the maximum extent possible, unique or proprietary elements of Krueger's chairs were removed to maximize competition." I do not, however, find Wallis' statement contrary to my conclusion about the inherent distinctiveness of the KI chair because although Unicor's specifications do include a drawing of the KI chair as an example, the specifications themselves do not call for any of the distinctive trade dress aspects of the KI chairs. For example, the specifications specify a general size for the backs and seats but do not require a particular shape. Similarly, nowhere do the specifications require a zig-zag, in-out shape for the connection between the back and seat.

Turning to the issue of functionality, defendant argues that the Matrix chair design is functional because only a clone of the Matrix chair will "interstack" perfectly with it. I disagree. As defendant should be well aware, the argument of perfect compatibility has not been embraced in this circuit for many years, and it was explicitly rejected in Wallace, 916 F.2d at 80 (rejecting Pagliero v. Wallace China Co., 198 F.2d 339 (9th Cir. 1952)). As a factual matter, neither Unicor nor the rest of the market has ever demanded perfect "interstackability." Unicor clearly explained during the bid process that interstackability was only one of many considerations. Although defendant's success in winning the contract may have had something to do with the fact that it had produced an exact replica of the Matrix chair, Unicor gave its second-highest ranking to competitor CPSI, whose chair looks nothing like the Matrix chair and does not stack perfectly with it. (The CPSI stacks adequately with the Matrix chair, but not nearly as well as the Matrix chair itself, which was ranked third in the bidding process.) Defendant's argument is further undercut by the fact that the Unicor contract is just one small segment of the market. There is no evidence that convention halls, school systems and other potential customers are obsessed with perfect interstackability among competing models. Indeed, as defendant admitted at oral argument, no chair currently on the market interstacks ideally with any other. (Tr. of Hearing at 44.)

To determine whether the design features at issue are functional, I must ask whether their shape is "dictated by the functions to be performed." Warner Bros., 724 F.2d at 331. The answer is clearly no. The function of a chair is to provide a place to sit. A stacking chair, in addition, must stack. On the evidence before me, many other designs serve these functions equally well. Therefore I find that the design features at issue are ornamental, not functional. Next I must address whether there are sufficient alternative arrangements to permit competition. Stormy Clime, 809 F.2d at 977. I find that defendant has not met its burden of showing that the supply of alternative designs would be unduly limited if defendant could not replicate the Matrix chair. See LeSportsac, 754 F.2d at 76 (burden falls on defendant to prove functionality). At least a dozen designs compete in this market, and one can readily imagine other hypothetical designs. Although the number of alternative designs is not limitless, given the need for stackability, the alternatives are not unduly limited so long as KI's trade dress is confined to the "precise expression of a decorative style." Wallace, 916 F.2d at 81.

I conclude, however, that KI may not claim a trade dress in each of its design features individually. The shape of its seat, in particular, is not highly distinctive in this market and its "dip" may have been designed largely for comfort, a functional consideration.⁶ KI has made a case, however, that the combination of the seat, the back and the Z-shaped connecting rods give the Matrix chair an aesthetically unique appearance. Because other competitors seem to have no trouble producing visually distinguishable chairs, defendant should be expected to do the same. Defendant need only change the shape of its chair back or seat sections, or avoid using Z-shaped connecting rods. Although changing the shape of the Beetle chair shape might make it imperfectly stackable with the Matrix chair, as discussed supra, neither Unicor nor the market generally requires such perfection.

6 At oral argument, defendant unveiled (practically with a drum-roll and a flourish) a specimen of the competing Staxx chair, whose seat is extremely similar in shape to the Matrix seat, although not identical.

B. Likelihood of Confusion

Having found that the overall look of the Matrix chair is inherently distinctive, I next turn to whether buyers are likely to be confused by the similarity between the two designs. As with the first prong of the test for trade dress infringement, the burden of proving likelihood of confusion falls on the plaintiff. In this circuit, the likelihood of confusion is weighed according to the eight factors set out by Judge Friendly in Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492, 495 (2d Cir.), cert. denied, 368 U.S. 820, 82 S.Ct. 36, 7 L.Ed.2d 25 (1961). The factors are: (1) the strength of the mark; (2) the degree of similarity between the two marks; (3) the proximity of the products; (4) the likelihood that the senior user will "bridge the gap" and enter the junior user's market; (5) actual confusion; (6) the defendant's good faith; (7) the quality of the defendant's product; and (8) the sophistication of the buyers. In addition, the Second Circuit has suggested other factors to be considered, if relevant, including the senior user's delay in asserting its claim. Centaur Communications, Ltd. v. A/S/M Communications, Inc., 830 F.2d 1217, 1228 n. 2 (2d Cir. 1987). The weighing of these factors is not a "mechanical process." Rather, "a court should focus on the ultimate question of whether consumers are likely to be confused." Paddington, 996 F.2d at 584 (citations omitted).

On the evidence before me, I conclude that plaintiff has shown a likelihood of confusion, but not such an overwhelming likelihood that a preliminary injunction is justified in light of its delay in bringing this action. Because the products compete in the same market, I need not conduct an extended analysis of proximity and bridging the gap, both of which strongly favor the plaintiff. I will address in some detail, however, the evidence relevant to the other factors.

1. Strength of the Trade Dress

A mark's strength is its "distinctiveness . . ., or more precisely, its tendency to identify the goods sold under the mark as emanating from a particular, although possibly anonymous, source." Paddington, 966 F.2d at 585 (citation omitted). Such strength may result from the mark's inherent distinctiveness, its secondary meaning, or both. Id.

For the purposes of this motion I have found the Matrix design to be inherently distinctive, but KI has not persuaded me that the design has acquired secondary meaning. The evidence required to prove secondary meaning includes advertising expenditures; consumer surveys showing that customers associate the trade dress with the source; sales success; unsolicited media coverage of the product; attempts to plagiarize the mark; and length and exclusivity of the mark's use. See, e.g., Thompson Medical, 753 F.2d at 217. The Matrix chair clearly meets the length and exclusivity test, having been on the market for 20 years, and defendant has recently tried to copy the mark, but the other evidence is far from dispositive.

Plaintiff's advertising expenditures have been surprisingly low. Over the last 20 years, KI has spent only about $600,000 to promote the Matrix chair, primarily through catalogs. In the last four years, KI has spent less than $40,000 per year on advertising. (Jaraczewski Aff. at 2, 4.) The nature of this advertising, moreover, does not encourage buyers to associate the Matrix design with KI. There are no slogans of the "look for" type, such as, "Look for the Z," or "It's not a KI without a zig-zag." Cf. Coach Leatherware, 933 F.2d at 168 (citing handbag advertising slogan: "It's Not a Coach Bag Without the Coach Tag"). The catalogues simply present photographs of the chair. On the other hand, design can serve as a source identifier in this industry, and many chairs are covered by design patents. Knock-offs are not an industry tradition, as they are with handbags or clothing. And KI hardly needed to advertise the uniqueness of a design that was protected by a design patent. There is plenty of room here for more evidence at trial, but the facts provided here for secondary meaning are inconclusive.

In place of a consumer survey, plaintiff provides four affidavits of modest value. Two affiants are employees of KI, who state that the Matrix design is immediately recognizable as a KI product. These are conclusory statements made by interested parties and thus carry little weight. The other affidavits are equally unhelpful. One affiant, an executive with the Las Vegas Convention and Visitors Authority, asserts that the Matrix is "unmistakable in its design" and that the convention center bought 12,000 Matrix chairs because of their "distinctive styling, user comfort and ease of handling." (Smith Aff. at 2.) This merely establishes that anyone who buys 12,000 chairs will make it his business to learn where they come from. The fourth affiant, a manager of a competing office furniture manufacturing company, states that he "immediately" recognizes the design as originating with KI. (Hieftje Aff. at 2.) This merely proves that competitors in this industry know each other's chairs. These affidavits do not establish secondary meaning.

KI's evidence of sales success is too sketchy to be persuasive. KI asserts that it has sold 3.7 million Matrix chairs in the last 20 years, including $150 million in U.S. sales. (Jaraczewski Aff. at 4.) At oral argument, KI's counsel added that of all the stacking chairs now in the hands of consumers, 25% are Matrix chairs. (Tr. of Hearing at 26-7.) KI has not, however, provided a sworn affidavit to this effect. Even if the Matrix chair dominates the industry, this merely proves that the design is popular, not that buyers associate the design with KI.

KI has offered no evidence of unsolicited media coverage.

Finally, although KI secured a registered trademark for Matrix, it has not sought to register its trade dress.

In sum, although I have found that the Matrix chair has a distinctive look, KI has not provided sufficient evidence of acquired secondary meaning. Therefore, I conclude that the mark is only moderately strong.

2. Similarity between the Marks

The chair designs are almost identical, but similarity alone does not create customer confusion. "Similarity in overall appearance alone cannot establish source confusion as a matter of law." Coach Leatherware, 933 F.2d at 169. Prominent labeling of highly similar products can dispel a likelihood of confusion. E.g., L.A. Gear, Inc. v. Thom McAn Shoe Co., 988 F.2d 1117, 1134 (Fed. Cir. 1993) (no confusion between similar athletic shoes when products prominently marked); Bristol-Myers Squibb Co. v. McNeil-P.P.C., Inc., 973 F.2d 1033, 1046 (2d Cir. 1992) (no confusion between "Excedrin PM" and "Tylenol PM" when labels serve as most prominent feature); Stormy Clime, 809 F.2d at 977 (no confusion between similar rain jackets when copier displays its name prominently on label, zipper pull, hang tag, and package); Bose Corp. v. Linear Design Labs, Inc., 467 F.2d 304, 309 (2d Cir. 1972) (no confusion between audio speakers when copier's name displayed on front of product).

When the labeling is subtle or obscure, however, the likelihood of confusion is higher. E.g., Lois Sportswear, U.S.A., Inc. v. Levi Strauss Co., 799 F.2d 867, 873-75 (2d Cir. 1986) (identical back-pocket stitching pattern likely to confuse buyers in post-sale context, when labels not visible); LeSportsac, 754 F.2d at 79 (strikingly similar designs, combined with inconspicuous and similar logos, likely to confuse buyers).

In this case, the KI's name is embossed into the plastic underneath the Matrix seat, an obscure place for a label. All Beetle chairs produced after mid-January will be similarly marked with NI's name. In addition, NI asserts that it will ship the Beetle chairs in prominently marked boxes. Such labeling may indeed prevent pre-sale confusion among architects and designers, who are the most sophisticated buyers in this market, but it will not help less sophisticated customers, who are unlikely to peer underneath the seat. As KI points out, such customers could well be confused post-sale. If they see a Beetle chair of inferior or different quality and assume it to be a Matrix chair, KI's reputation will suffer. Moreover, about 7,300 Beetle chairs will never be permanently labeled because they were manufactured before NI's embossing mechanism had been completed. (Def.'s Mem. in Opp. at 19.) Although NI promises that these unmarked chairs will be equipped with hang tags, this cannot be a complete solution. As this circuit has observed, hang tags are removable. "Once the hang tag comes off, the [products] are again virtually identical and likelihood of confusion remains." LeSportsac, 754 F.2d at 80. In NI's favor, most of the unmarked Beetle chairs are destined for Unicor, which is unlikely to be confused about the source. On balance, the similarity of the marks weighs strongly in plaintiff's favor.

3. Actual confusion

Evidence of actual confusion is not required, because actual confusion is difficult to prove and the law requires only likelihood of confusion. E.g., Lois Sportswear, 799 F.2d at 875. Although KI cites one incident of actual confusion, its probative value is minimal. The incident arose when one of its own sales representatives picked up three sample "Matrix" chairs from someone who had earlier solicited bids on the chairs and was ready to return the chair samples. A few days later, the customer informed the KI rep that he had picked up one Beetle chair by mistake. NI had won the bid, and Beetle samples had been mixed in with the Matrix samples. In an affidavit, the KI sales rep states that the Beetle chair "stacked perfectly on the authentic Matrix chairs and was impossible to distinguish from the authentic Matrix chairs even after a long examination." (Walsh Aff. at 2.) This evidence does not establish the kind of customer confusion that trade dress law was designed to prevent. There is no evidence that the customer was so confused, pre- or post-sale, that it mistakenly awarded the contract to NI. Any post-sale confusion was quickly remedied by the customer, who called to explain the mix-up. This anecdote proves little one way or the other, and must be evaluated in light of the sophistication of the customers, infra.

4. Defendant's good faith

Defendant admits that it copied the Matrix chair. Intentional copying gives rise to a presumption of likelihood of confusion. E.g., Perfect Fit Indus., Inc. v. Acme Quilting Co., 618 F.2d 950, 954 (2d Cir. 1980). Defendant has little excuse for such intentional copying, even if what Unicor really wanted was a low-priced copy of the Matrix chair. (This Court makes no determination of what Unicor really wanted.) Other bidders, including the second-ranked CPSI, managed to avoid copying the Matrix chair. Similarly, it was incumbent upon defendant to distinguish its product either by creating a unique design or by displaying its name more prominently. However, there is no evidence that defendant was trying to deceive customers into thinking they were buying a Matrix chair. Unicor, which bought the Beetle chairs through a competitive bidding process, was never in danger of being confused or deceived. Defendant is walking a fine line here between trading on KI's goodwill and openly advertising that it is selling a knock-off. Under the circumstances, this factor weighs against defendant, but only moderately.

5. Quality of Defendant's Product

The evidence on this score is inconclusive, raising only a possibility that defendant's product may be inferior. KI presents a three-page affidavit from one of its own engineers, who recently examined a Beetle chair and found numerous flaws. (Nordgren Aff. at 2-3.) Defendant counters with an affidavit from its president, William Breen, who asserts that KI must have tested a "preproduction prototype" and that the actual production run is of the highest quality. (Breen Decl. at 8.) A test on one chair, possibly from an early production run, is not persuasive evidence of inferior quality.

6. Sophistication of the Buyers

High density stacking chairs are sold to two tiers of customers, one highly sophisticated and the other moderately so. The highly sophisticated tier includes architects, designers and contract furniture representatives, who decorate and furnish convention centers and school systems through a competitive bidding process. (Tr. of Hearing at 11-12, 38). Chairs are often just a part of their total bid package. As plaintiff testified at oral argument, such architects and designers often supply large quantities of chairs: "It may be a hundred, it may be a thousand, it may be ten thousand if we are talking about a convention center." (Tr. at 12.) The second group of customers includes executives at the convention centers and the school systems themselves, who sometimes choose the chairs directly from among choices supplied by the architects or approve the chairs proposed by bidders.

The relative sophistication of this market may weigh in defendant's favor, but how much is difficult to tell. In general, the more sophisticated the customers, the less likely they are to be confused about the source of the product. See, e.g., Paddington, 996 F.2d at 587. However, even sophisticated buyers are not always careful buyers, and their very awareness of status brand names and designs may make them more vulnerable to confusion. E.g., Grotrian, Helfferich, Schulz, Th. Steinweg Nachf. v. Steinway Sons, 523 F.2d 1331, 1340 (2d Cir. 1975) (sophisticated buyers of quality pianos more likely to mistakenly associate "Grotrian-Steinweg" pianos with "Steinway" pianos). As the district court noted in Steinway, the "subliminal confusion" between similar marks or designs "can transcend the competence of even the most sophisticated customer." Id. at 1341. See also Lois Sportswear, 799 F.2d at 875 ("Presumably it is these sophisticated jeans buyers who pay the most attention to back pocket stitching patterns and their `meanings.'") In the instant case, there has been no testimony on how the products are priced, an important factor in this calculus. Sophisticated architects and designers are likely to know when new products enter the market, even if the new products are similar to established products. Such buyers are not likely to be confused when a copy is significantly less expensive, which may be the case here. The same logic applies if the copy is of grossly inferior quality, a problem KI contends is present here. But as noted above, it is possible that the managers of convention centers (and similar customers) might mistake one chair for the other when choosing from catalogues or approving chairs, with resulting harm to KI.

7. Senior User's Delay in Asserting Its Claim

This factor weighs strongly against the plaintiff, despite defendant's brazen act of deliberate copying. One of plaintiff's own executives saw the Beetle prototype in March 1995, nine months before this action was filed. Plaintiff's only response to seeing the prototype was to protest that defendant was using KI parts. Plaintiff expressed no alarm that defendant had already duplicated the distinctive Matrix frame, and plaintiff did not assert or warn about any trade dress rights at that time. Plaintiff argues that a mere prototype does not necessarily indicate the final design, that defendant falsely represented that the prototype was only a general illustration of the product, and that plaintiff should not be required to file premature lawsuits. These purported justifications are not persuasive. KI has consistently maintained throughout this action that the single most distinctive element of its trade dress is the Z-shaped connecting rods. The Beetle prototype frame, made by defendant, featured precisely those Z-shaped rods. The Matrix seat and back fit defendant's frame with perfection. KI's lack of suspicion and further active investigation at this point is baffling. Moreover, sometime in or after June 1995, the same KI executive who had seen the Beetle prototype discovered defendant's new catalogue in the Unicor offices and "noticed the photographs of the Beetle chair." (Campanale Decl. at 2.) And still KI voiced no concern.

It is unclear what KI's attorney should or should not have done in May 1995, when he received a photograph of the Beetle chair among other materials. If the photograph was covered by the GAO protective order, KI's counsel still could have asked the GAO for a waiver of the protective order or alerted defendant's counsel to the possibility of a trade dress claim. This is a moot point, however, because defendant's public catalogue was in or should have been in KI's hands in or around June 1995, just a month later when NI began shipping the catalogues to Unicor and its dealers and representatives.

Another six months elapsed before KI took action. Such nonchalance is inexplicable. KI passed up many opportunities to assert its rights, such as sending a cease and desist letter to defendant, alerting Unicor that KI had proprietary rights in the Matrix chair design, and launching a prompt investigation either after it saw the prototype in March 1995, or even after seeing the chair at the first trade show on November 2, 1995. The evidence suggests that plaintiff did not begin to think about a trade dress infringement claim until it both lost the Unicor contract and the Beetle chair began appearing at trade shows — i.e., until KI decided to take the competitive threat seriously. The trade dress laws, however, were not designed to protect producers from competitive injury.

8. Balancing the Factors

Having considered the nine factors relevant to the likelihood of confusion, I find that three factors weigh heavily in plaintiff's favor: similarity, proximity, and bridging the gap. Two additional factors, the strength of the mark and defendant's good faith, weigh moderately in plaintiff's favor. Three factors favor neither side: evidence of actual confusion, the quality of defendant's product, and the sophistication of the buyers. The ninth factor, delay, undercuts much of plaintiff's argument. Had KI asserted a forceful trade dress claim earlier — say, in the summer of 1995 when it first saw or should have seen a photograph of the Beetle chair, and when defendant's production line was idled by the bid protest — defendant might have been persuaded to change its design. Plaintiff's silence allowed defendant and Unicor to believe that no trade dress claim would be forthcoming. This tips the equities more in defendant's favor than the other factors would otherwise indicate. Thus, although plaintiff has not met its burden for preliminary relief, plaintiff has established a likelihood of confusion sufficient to go to trial.

II. Motion for Preliminary Injunction

In order to obtain a preliminary injunction, the moving party must show (1) the likelihood of irreparable injury, and (2) either (a) likelihood of success on the merits, or (b) sufficiently serious questions going to the merits and a balance of hardships tipping decidedly in the movant's favor. Tough Traveler, 60 F.3d at 967. Ordinarily, a likelihood of confusion establishes a presumption of irreparable harm. Id. This presumption does not apply, however, when a plaintiff has delayed without good cause in seeking emergency relief. As this circuit has observed:

Preliminary injunctions are generally granted under the theory that there is an urgent need for speedy action to protect the plaintiffs' rights. Delay in seeking enforcement of those rights, however, tends to indicate at least a reduced need for such drastic, speedy action. . . . Significant delay in applying for injunctive relief in a trademark case tends to neutralize any presumption that infringement alone will cause irreparable harm pending trial, and such delay alone may justify denial of a preliminary injunction for trademark infringement.

Citibank, N.A. v. Citytrust, 756 F.2d 273, 276 (2d Cir. 1985) (no presumption of irreparable harm when plaintiff delayed suit for nine months after receiving notice of alleged infringement in the press, and 10 weeks after receiving actual notice).

In the instant case, for the reasons discussed, I decline to manufacture a sense of urgency that is not supported by plaintiff's own conduct. Although the law urges leniency toward a plaintiff who has good cause for delay, either because the plaintiff did not know how severe the infringement was or because the plaintiff was making good faith efforts to investigate, neither of these justifications applies here. See Tom Doherty Assocs., Inc. v. Saban Entertainment, Inc., 60 F.3d 27, 39 (2d Cir. 1995) (citations omitted). The courts have often denied preliminary injunctions to plaintiffs whose delay suggested that irreparable harm was unlikely. E.g., Tough Traveler, 60 F.3d at 968 (nine month wait before filing complaint is too long, particularly when plaintiff delayed another four months before moving for preliminary injunction); Century Time Ltd. v. Interchron Ltd., 729 F. Supp. 366 (S.D.N Y 1990) (six month wait is too long); Comic Strip, Inc. v. Fox Television Stations, Inc., 710 F. Supp. 976 (S.D.N.Y. 1989) (seven months is too long); Nina Ricci, S.A.R.L. v. Gemcraft Ltd., 612 F. Supp. 1520 (S.D.N.Y. 1985) (four months, including six weeks from breakdown of settlement talks, is too long). In Doherty, the Second Circuit observed that delay rebuts the presumption of irreparable harm when one can draw a "fair inference" that "the owner of the mark or right had concluded that there was no infringement but later brought an action because of the strength of the commercial competition." Doherty, 60 F.3d at 39. KI's course of conduct before Unicor and its lack of careful monitoring of NI's activities rebut its claim of irreparable injury.

On balance, although I have found that some likelihood of confusion exists, I have also found that this likelihood is not overwhelming because of the general sophistication of the buyers in this market, particularly of Unicor, the biggest current buyer of NI's chair. Plaintiff's delay in bringing this action tips the equities in favor of defendant, which will suffer considerable economic harm if forced to cancel its contract with Unicor or to immediately re-tool to alter its design for the Unicor contract. Unicor, whose prison facilities depend on the supply on product, will also be unfairly prejudiced in the delay in receiving products occasioned solely by the late assertion by KI of its trade dress rights.

I remind defendant, however, that KI has raised serious questions of trade dress infringement, that this case is destined for trial, and that the requirements for a finding of success on the merits are significantly lower than the preliminary injunction standard at issue here.

CONCLUSION

In summary, the plaintiff has shown sufficiently serious questions going to the but has not shown a likelihood of irreparable harm or a balance of hardship tipping decidedly in its favor given its delay in asserting its rights. Plaintiff's motion for emergency relief came six to nine months after plaintiff learned of defendant's plans to copy its Matrix chair. This is too late. For the foregoing reasons, plaintiff's motion for a preliminary injunction is denied.

SO ORDERED.