John H. Harland Company, Cross-Appellant v. Clarke Checks, Inc., Cross-Appellee

R. LANIER ANDERSON, III, Circuit Judge: In this case, we are required to enter once again “the rather swampy area of unfair competition.” B.H. Bunn Co. v. AAA Replacement Parts Co., 451 F.2d 1254 (5th Cir.1971) (Goldberg, J.). The court below dismissed appellee’s claim for copyright infringement, but awarded damages and granted injunctive relief for trademark infringement, trade dress infringement, and a variety of state law claims. After wading through a morass of exhibits, briefs and record volumes, we affirm the judgments rendered below but remand so that the trial court can mark out a clearer path for appellants to follow through the slough of unfair competition law. I. FACTS John H. Harland Co. (“Harland”), a Georgia corporation with headquarters in Atlanta, is a major manufacturer and seller of bank stationery products, including personalized bank checks. Clarke Checks, Inc. (“Clarke”), a Texas corporation with headquarters in San Antonio, also manufactures and sells bank stationery products, including personalized bank checks. Unlike Har-land, which markets its products nationally, Clarke’s business is primarily based in the south and southeastern portions of the United States. Both companies sell their products through catalogs and brochures which are shown to consumers by new accounts personnel at banking institutions. Harland and Clarke, as well as other bank stationery companies, have long manufactured three-checks-to-a-page desk-style checkbooks. In these checkbooks, the checks typically are attached to permanent check stubs which can be used to record check transactions. A perforation makes it easy to detach the check from the stub, which remains in the checkbook as a permanent record. Such desk-style checkbooks are relatively large and, unlike wallet-style checkbooks, not readily portable. Although a person can remove one or more checks from the checkbook and carry them to other locations, the checkwriter does not have either a stub or a ledger on which to record the transaction. Thus, this type of checkbook traditionally has been used primarily for business accounts rather than personal accounts. In 1976, Harland began marketing a new product which solves the portability problem commonly associated with three-checks-to-a-page desk-style checkbooks. Harland inserted a small, carry-around stub between the check and the checkbook’s permanent stub. This intermediate stub has lines for recording the date, the amount of the check, the payee, and the purpose of the check, the same information which normally would be placed on a permanent stub. The addition of this intermediate stub makes it more convenient to use checks from the desk-style checkbook at various locations. A person can remove the checks and the corresponding carry-around stubs from the checkbook, 1 placing them in a carrying case which comes with the checkbook; then, when each check is written, the check writer can record the transaction on the carry-around stub, detach the stub, place it in the carrying case, and keep the stub until there is an opportunity to enter the transaction onto the conventional check stub in the desk-style checkbook. *970 Harland spent considerable time and effort marketing its new product, heralding the intermediate stub as “the most functional innovation ever offered in three-to-the-page checks.” 2 Harland named the intermediate stub the Memory Stub and registered this name as a trademark with the U.S. Patent and Trademark Office on December 6, 1976. Harland also registered copyright claims with the U.S. Copyright Office and placed copyright notices on all of its intermediate stub products. 3 The new product as a whole, a three-checks-to-a-page desk-style checkbook with intermediate carry-around stubs and an accompanying carrying case with compartments for the checks and stubs, was marketed as the “Odyssey Collection” and promptly became one of Harland’s most successful products. In 1979, Clarke modified its standard three-checks-to-a-page desk-style checkbook to include an intermediate carry-around stub. There is no doubt from the record that Clarke’s modification was a response to the competitive success of Harland’s Odyssey Collection. Nor is there any doubt that Clarke’s development process consisted largely of copying Harland’s product. Clarke officials reviewed Harland’s catalog and advertising materials, placed an order for some of Harland’s Odyssey Collection cheek packages, and referred to Harland’s products when designing Clarke’s new desk-style format. 4 Not surprisingly, Clarke’s new product was very similar to Harland’s: both were three-checks-to-a-page desk-style checkbooks with intermediate carry-around stubs; the intermediate stubs were identical in size and shape; both intermediate stubs had vertical lines for recording the same information; and the products had virtually identical cases for carrying the checks and intermediate stubs. See Figures 3-6. Clarke marketed its product under the name Deskette Entry Stub and called the intermediate stub an Entry Stub. When Clarke began marketing this new product, Harland informed Clarke that it was engaging in “infringing acts” and demanded that Clarke withdraw the product from the market. Clarke refused, and this lawsuit followed. In its complaint, Harland brought claims for trademark infringement, unfair competition (including trade dress infringement), trademark dilution, injury to business reputation, deceptive trade practices and copyright infringement, arising under the common law, the statutes of the State of Georgia, the Trademark Act of 1946 (commonly known as the Lanham Act), 15 U.S.C.A. *971 §§ 1051 et seq., and the Copyright Act of 1976, 17 U.S.C.A. §§ 101 et seq. The district court dismissed Harland’s copyright infringement claim, but conducted a five-day jury trial on the other claims. The jury found Clarke liable for trademark infringement and trade dress infringement, 5 awarding Harland $50,000 in damages. The district court then entered judgment against Clarke on all claims except the copyright claim and, inter alia, enjoined Clarke from using “trade dress for bank check products which is confusingly similar to the trade dress or overall appearance” of Harland’s Memory Stub check products. On appeal, appellee/cross-appellant Har-land contends that the district court erred when it granted summary judgment in favor of Clarke on the copyright infringement claim. Appellant/cross-appellee Clarke contends that the evidence was insufficient to support the findings of trademark infringement and trade dress infringement. Clarke also contends that the district court’s injunction is overbroad and does not set forth in sufficient detail the act or acts sought to be restrained. We consider each of these contentions below. II. COPYRIGHT INFRINGEMENT The first issue on appeal concerns the propriety of the district court’s granting of Clarke’s motion for summary judgment on the copyright infringement claim. The district court held that Harland’s Memory Stub product was not subject to copyright protection because it was merely a blank form which did not convey information. Thus, the district court concluded that Har-land’s copyright registrations only protected the artistic background designs on the Odyssey Collection checks. 6 Harland contends that this conclusion was erroneous. According to Harland, the copyright registrations not only protected the background designs on the checks, but also protected the entire “Memory Stub expression” which “consists of the specific size, shape, layout and decoration of the earry-around stub and its placement in relation to other portions of the check product.” Brief of Appel-lee/Cross-Appellant at 36-37. 7 We agree with the district court that Harland’s Memory Stub product was not copyrightable. Consequently, we affirm the district court’s grant of summary judgment on this issue. It is well-established that blank forms which do not convey information or contain original pictorial expression are not copyrightable. See, e.g., Baker v. Selden, 101 U.S. 99, 107, 25 L.Ed. 841 (1879) (holding that “blank account-books are not the subject of copyright”); M.M. Business Forms Corp. v. Uarco, Inc., 472 F.2d 1137, 1139 (6th Cir.1973) (“Generally, forms, including blank forms, which are intended to be used for recording facts are not the proper subjects of copyright.”); Time-Saver Check, Inc. v. Deluxe Check Printers, Inc., 178 U.S.P.Q. (BNA) 510 (N.D.Tex.1973) (holding that blank bank checks and attached duplicate forms to be used with carbon paper for record purposes are not copyrightable). This rule is embodied in regulations on the scope of copyright protection promulgated by the Copyright Office, which state in relevant part: The following are examples of works not subject to copyright and applications for *972 registration of such works cannot be entertained: (c) Blank forms, such as time cards, graph paper, account books, diaries, bank checks, score cards, address books, report forms, order forms, and the like, which are designed for recording information and do not in themselves convey information .... 37 C.F.R. § 202.1 (1982) (emphasis added). 8 Harland’s Memory Stub does not convey any information beyond that contained on previously existing check stubs which are manufactured by numerous bank stationery companies. The Memory Stub merely provides lines on which the check writer can record the date, the dollar amount of the check, the payee of the check, and the purpose of the check. Cf. Time-Saver Check, Inc. v. Deluxe Check Printers, Inc., supra. Thus, we agree with the district court that Harland’s Memory Stub is merely designed for recording information and does not convey information or contain original pictorial expression. Consequently, we affirm the grant of summary judgment on the ground that the “Memory Stub expression” was not copyrightable. III. TRADEMARK INFRINGEMENT Clarke contends that the district court erred when it declined to grant either a directed verdict or judgment notwithstanding the verdict on Harland’s trademark infringement claim. According to Clarke, the jury should not have been permitted to consider the trademark infringement issue because, as a matter of law, Clarke’s Entry Stub mark does not infringe on Harland’s registered Memory Stub mark. Section 32(1) of the Lanham Act, 15 U.S.C.A. § 1114(1), governs lawsuits for the infringement of a federally registered trademark. A defendant is liable for infringement if, without the consent of the registrant, he uses “in commerce any reproduction, counterfeit, copy, or colorable imitation of a registered mark” which “is likely to cause confusion, or to cause mistake, or to deceive.” Id. Thus, the critical question in most actions under § 32(1) is whether there is a likelihood of confusion, mistake, or deception between the registered mark and the allegedly infringing mark. See, e.g., Sun-Fun Products, Inc. v. Suntan Research & Development, Inc., 656 F.2d 186, 189 (5th Cir.1981). 9 This court and the former Fifth Circuit have identified a number of factors which should be considered when analyzing whether there is a likelihood of confusion between two marks. Among the factors are “the type of trademark, the similarity of design, the similarity of the product, the identity of retail outlets and purchasers, the similarity of advertising media used, defendant’s intent, and actual confusion.” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d 500, 504 (5th Cir.1980); see also Original Appalachian Artworks, Inc. v. Toy Loft, Inc., 684 F.2d 821, 832-33 (11th Cir.1982) (applying essentially the same factors to determine whether there was likelihood of confusion in a trade dress infringement case); Safeway Stores, Inc. v. Safe *973 way Discount Drugs, Inc., 675 F.2d 1160 (11th Cir.1982) (applying same factors in § 32(1) case involving registered service mark). In this case, the district court properly instructed the jury on the factors it should consider in determining whether there was a likelihood of confusion between Harland’s registered Memory Stub trademark and Clarke’s use of the mark Entry Stub. 10 The jury necessarily found that there was a likelihood of confusion between the two marks because it returned a verdict against Clarke for trademark infringement. The jury’s finding on the likelihood of confusion issue is factual, Original Appalachian Artworks, Inc. v. Toy Loft, Inc., 684 F.2d at 832; Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., 675 F.2d at 1163; Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d 252, 258 (5th Cir.), cert. denied, 449 U.S. 899, 101 S.Ct. 268, 66 L.Ed.2d 129 (1980), and must be affirmed if it was based on substantial evidence, i.e. “evidence of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment might reach different conclusions.” Boeing Co. v. Shipman, 411 F.2d 365, 374 (5th Cir. 1969). 11 Although the issue is a close one, we have concluded that the jury’s verdict on the trademark infringement issue must be affirmed. An evaluation of the factors which this circuit has identified as relevant in trademark infringement cases convinces us that a reasonable and fair-minded jury could find that there was a likelihood of confusion between Harland’s registered Memory Stub trademark and Clarke’s use of the Entry Stub mark. A. Type of Trademark. The first factor to consider is whether Memory Stub is a “strong” or a “weak” trademark. “[T]he strength and distinctiveness of plaintiff’s mark is a vital consideration in determining the scope of protection it should be accorded. ‘Strong marks are widely protected, as contrasted to weak marks.’ ” Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 259 (quoting Lunsford, Trademark Basics, 59 Trade-Mark Rep. 873, 878 (1969)). Harland contends that Memory Stub is a strong trademark. To support this contention, Harland focuses on several former Fifth Circuit cases which indicated that “[a] strong trademark is one that is rarely used *974 by parties other than the owner of the trademark, while a weak trademark is one that is often used by other parties.” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d at 504; see also Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 259-60 (evidence of extensive third-party uses of “Domino” mark limited the scope of the protection to be accorded plaintiff’s mark); Armstrong Cork Co. v. World Carpets, Inc., 597 F.2d 496, 505 (5th Cir.) (extensive third-party uses of the mark “World” suggested that there was no infringement), cert. denied, 444 U.S. 932,100 S.Ct. 277, 62 L.Ed.2d 190 (1979); Restatement of Torts § 729, Comment (g), at 596 (1938) (“The greater the number of identical or more or less similar trade-marks already in use ... the less is the likelihood of confusion.”). Harland points out that there is no evidence that Memory Stub or any similar mark has been used on any other products by third parties. Accordingly, Harland argues, Memory Stub is a strong trademark which deserves broad protection. Clarke disagrees with Harland’s characterization of Memory Stub as a strong mark and with Harland’s method of determining the “strength” of the mark. According to Clarke, the key consideration in determining the strength of a trademark is whether the mark is “arbitrary” or “fanciful,” “suggestive,” or merely “descriptive.” An arbitrary or fanciful mark has no inherent relationship to the product or service with which it is associated. A suggestive mark suggests some characteristic of the product or service to which it is applied, but requires the consumer to use his imagination to determine the nature of the product or service. A descriptive mark describes a characteristic or quality of the product or service, such as its intended use, its ingredients, its dimensions, its desirable features, or its end effect on the consumer. See Sun Banks v. Sun Federal Savings & Loan Ass’n, 651 F.2d 811, 315-16 (5th Cir.1981) (discussing classification of service marks and trademarks); Soweco, Inc. v. Shell Oil Co., 617 F.2d 1178, 1183-84 (5th Cir.1980) (discussing categories of trademarks), cert. denied, 450 U.S. 981, 101 S.Ct. 1516, 67 L.Ed.2d 816 (1981). A purely fanciful or arbitrary mark is generally considered strong and is given “protection over a wide range of related products and variations in appearance of the mark.” 1 J.T. McCarthy, Trademarks & Unfair Competition § 11:24, at 398 (1973). A descriptive mark, on the other hand, is considered weak and is given a “narrow range of protection.” Id. A suggestive mark falls somewhere between the two preceding types of marks and, although a suggestive mark can be protected without evidence that it has acquired secondary meaning, 12 is comparatively weak. See, e.g., Sun Banks v. Sun Federal Savings & Loan Ass’n, 651 F.2d at 315. Clarke contends that Memory Stub is a weak mark because it is “at least suggestive if not merely descriptive.” Brief of Appellant/Cross-Appellee at 35. 13 Consequently, *975 Clarke argues, Harland’s trademark is not entitled to broad protection. Recent authority in this circuit indicates that both of the factors cited by the parties should be considered when analyzing the strength of a particular trademark. In Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., supra, for example, the court discussed the categories of marks and then stated: “The words of a mark are not, however, the sole factors determining its strength. Also important is the extent of third-party use of the mark.” 675 F.2d at 1164. Cf. Sun Banks v. Sun Federal Savings & Loan Ass’n, 651 F.2d at 315 (“The ultimate strength of a mark ... is determined by a number of factors which establish its standing in the marketplace.”); Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 259-60 (discussing both number of third-party uses and categorization of mark in concluding that “Domino” was entitled to only limited protection outside the product areas in which plaintiff already had used its mark); 1 J.T. McCarthy, supra, § 11:24, at 401 (suggesting that the “ ‘strong-weak’ classification should be based both upon the amount of use of the term by others in this product and geographical area and upon the quantum of ‘strength’ and consumer recognition of the mark at issue”). Thus, while the fact that Memory Stub and similar marks have not been used by other parties normally might indicate that Harland’s mark is strong, this indication is significantly weakened by the suggestive or possibly even descriptive nature of the Memory Stub mark. Consequently, we do not believe that Harland’s Memory Stub mark is a particularly strong mark. This does not mean, of course, that the Memory Stub mark cannot be infringed upon. See AMF, Inc. v. Sleekcraft Boats, 599 F.2d 341, 350 (9th Cir.1979) (although a weak mark is entitled to only a “restricted range of protection,” infringement will be found “if the marks are quite similar, and the goods closely related .... ”); New West Corp. v. NYM Co., 595 F.2d 1194, 1202 (9th Cir.1979) (even if appellee’s mark was weak, appellee was entitled to protection “due to the obvious similarities of the marks in appearance and sound, along with the identical product and market distribution area”). Accordingly, we must carefully examine the remaining factors in order to determine whether, despite the fact that Harland’s mark is not particularly strong, a reasonable jury could nonetheless conclude that Clarke’s Entry Stub mark creates a likelihood of confusion between the two marks. B. Similarity of Design. “The similarity of design is determined by considering the overall impression created by the mark[s] as a whole rather than simply comparing individual features of the marks.” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d at 504-05. This consideration includes a comparison of the appearance, sound and meaning of the marks, as well as the manner in which the marks are used. See 2 J.T. McCarthy, supra, §§ 23:4, 5, 7, 8, 15, 18; Restatement of Torts § 729(a) (1938); Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 260-61; Armstrong Cork Co. v. World Carpets, Inc., 597 F.2d at 501-04. There are some obvious similarities between the appearance, sound and meaning of the mark Memory Stub and the mark Entry Stub. Both marks consist of a relatively short word which ends with “ry” followed by the word “Stub.” Cf. AMF, Inc. v. Sleekcraft Boats, 599 F.2d at 351 (Slickcraft/Sleekcraft); National Association of Blue Shield Plans v. United Bankers Life Insurance Co., 362 F.2d 374 (5th Cir. 1966) (Blue Shield/Red Shield). In the context of bank check products, however, *976 “stub” is a generic word which anyone is free to use. Cf. Beech-Nut, Inc., v. Warner-Lambert Co., 346 F.Supp. 547, 549 (S.D.N.Y. 1972) (holding that “Breath Pleasers” did not infringe on “Breath Savers” in part because “where the key word [breath] is as free as air, small variations are likely to make enough of a difference to ward off charges of infringement”), aff’d, 480 F.2d 801 (1973). Thus, although the marks ultimately must be considered as a whole, the focus of the inquiry regarding appearance and sound is on the non-generic portion, see 2 J.T. McCarthy, supra, § 23:15(G), and while the appearance of the word “Entry” is readily distinguishable from that of the word “Memory,” the sound and cadence of the two words are somewhat similar. With respect to meaning, there is a clear connection between the meaning of Memory Stub and Entry Stub as used on the parties’ products, 14 but there also is a distinction between the connotation of Memory Stub, which suggests the process of retaining and/or recalling information, and Entry Stub, which suggests making a record of a transaction for bookkeeping purposes. Although the similarities in appearance, sound and meaning between the marks Memory Stub and Entry Stub are far from overwhelming, these similarities are accentuated by the manner in which the marks are used. Clarke, like Harland, places its mark on the intermediate carry-around stub. See Figures 3 and 4. Although Clarke’s mark is in smaller print and is located at the bottom of the stub rather than at the top, both marks are placed within the decorative information box on the stub and both are used to label or “name” the stub. Further, both parties use the marks in their advertising materials and in the catalogs which they make available to new accounts personnel at banks. Thus, we believe that there are some similarities in the overall commercial impression created by Harland’s Memory Stub mark and Clarke’s Entry Stub mark. 15 C. Similarity of Products. “The greater the similarity between the products and services, the greater the likelihood of confusion.” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d at 505. In this case, the products are virtually identical. D. Identity of Retail Outlets and Purchasers. “Dissimilarities between the retail outlets for and the predominant customers of plaintiff’s and defendant’s goods lessen the possibility of confusion, mistake, or deception.” Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 262. Harland and Clarke sell their products through the same retail outlets— banking institutions — and compete for exactly the same customers in the south and southeastern portions of the United States. E. Similarity of Advertising Media Used. “The fifth factor used in evaluating the likelihood of confusion is the similarity between the parties’ advertising campaigns. The greater the simiarity in the campaigns, the greater the likelihood of confusion.” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d at 506. In this case, Harland and Clarke use virtually identical methods of advertising their products, relying primarily on catalogs which are available at banks. Both parties also use, to a lesser *977 extent, advertising brochures and leaflets as well as ads in trade journals. F. Defendant’s Intent. Although objective factors are most important in assessing the likelihood of confusion between two marks, see 2 J.T. McCarthy, supra, § 23:30, courts also examine the defendant’s subjective intent. Indeed, the former Fifth Circuit once stated that if “a plaintiff can show that a defendant adopted a mark with the intent of deriving benefit from the reputation of the plaintiff, that fact alone ‘may be sufficient to justify the inference that there is confusing similarity.’ ” Exxon Corp. v. Texas Motor Exchange, Inc., 628 F.2d at 506 (quoting Restatement of Torts § 729, Comment (f) (1938)); see also Sun Banks v. Sun Federal Savings & Loan Ass’n, 651 F.2d at 318-19 (“That a latecomer adopts another’s name or mark, deliberately seeking to capitalize on the other’s reputation and benefit from the confusion, is an important factor for any court.”). In this case, there is some evidence from which the jury might have inferred that Clarke adopted the Entry Stub mark in order to appropriate some of the good will associated with Harland’s Memory Stub mark. Clarke was well aware of the success of Harland’s Memory Stub product. More importantly, the record suggests that the officials who developed Clarke’s Entry Stub product attempted to copy Harland’s product as closely as possible, even using one of Harland’s Memory Stub products as a model when “designing” the new Clarke product. 16 Thus, although there is no direct evidence of Clarke’s intent, we believe the jury might have inferred that Clarke purposely chose a mark which was very similar to Memory Stub in order to benefit from the reputation Harland’s mark had already achieved. 17 Clarke argues that any possible inference of intent to trade on the reputation of Harland’s mark by adopting a confusingly similar mark is necessarily negated by evidence that Clarke officials consulted trademark counsel before adopting the Entry Stub mark. 18 We disagree. Although the evidence of Clarke’s consultations with counsel can be read to suggest that Clarke had no intent to adopt a mark which was confusingly similar to Harland’s mark, the jury also might have inferred that Clarke merely was attempting to come as close to Harland’s mark as the law would allow. Cf. Chevron Chemical Co. v. Voluntary Purchasing Groups, Inc., 659 F.2d 695, 704 (5th Cir.1981) (Unit A) (district court found that defendant “meant to copy [plaintiff’s] trade dress, ‘at least as much as the law would allow’ ”; former Fifth Circuit stated that the only plausible explanation for such behavior was that defendant had intent to *978 “cash in” on plaintiff’s good will), cert. denied, 457 U.S. 1126, 102 S.Ct. 2947, 73 L.Ed.2d 1342 (1982). Thus, the fact that Clarke officials sought legal advice before adopting the Entry Stub mark does not preclude a finding that Clarke intended to trade on the good will associated with Harland’s mark by adopting a mark which “approach[ed] so near to [the Memory Stub mark] that the public may fail to distinguish between them.” Florence Manufacturing Co. v. J.C. Dowd & Co., 178 F. 73, 75 (2d Cir.1910). Cf. Wolfe v. National Lead Co., 272 F.2d 867, 871 (9th Cir.1959) (fact that appellant acted on advice of counsel, did not preclude an accounting of profits because “[w]hether he believed himself to be within the law or not, he was knowingly and deliberately cashing in upon the good will of appellee”), cert. denied, 362 U.S. 950, 80 S.Ct. 860, 4 L.Ed.2d 868 (1960). G. Actual Confusion. “Although evidence of actual confusion is not necessary to a finding of likelihood of confusion, it is nevertheless the best evidence of likelihood of confusion.” Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d at 263. See also Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., 675 F.2d at 1166-67 (discussing the significance of evidence of actual confusion). Harland contends that it “introduced clear evidence of instances of actual confusion” at trial. Brief of Appellee/Cross Appellant at 14. In contrast, Clarke argues that “Harland attempted to show actual confusion, but utterly failed.” Brief of Appellant/Cross-Appellee at 37. Our own examination of the record indicates that although the evidence of actual confusion was not sufficient to compel a finding of likelihood of confusion, it was “worthy of some consideration.” Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., 675 F.2d at 1167 (holding that two instances of actual confusion, “one misdirected dunning letter from a creditor of [defendant] and one customer inquiry at [defendant],” were sufficient evidence of actual confusion to be “worthy of some consideration”); cf. World Carpets, Inc. v. Dick Littrell’s New World Carpets, 438 F.2d 482, 490 (5th Cir.1971) (noting that “reason tells us that ... very little proof of actual confusion would be necessary to prove the likelihood of confusion”). At trial, Harland presented evidence of two instances of actual confusion. Both instances involved order forms for intermediate carry-around stub products which were received by Clarke but which contained the word “Memory.” In one instance, the order form clearly asked for Clarke’s “3-to-page Deskette” with Clarke’s “World of Beauty” artistic background design, but a card attached to the order form included the words “Cover and Memory Case.” (Emphasis added.) 19 In the other instance, a bank new accounts representative checked the “Deskette 300” box on a Clarke order form and then added the handwritten notation “with memory stub.” (Emphasis added.) 20 Admittedly, neither *979 instance involved an intended order for Harland’s Memory Stub product which was misdirected to Clarke, 21 and neither instance indicated that bank customers, the ultimate consumers, were confused. 22 Nevertheless, we believe that the jury reasonably could have inferred that there was a likelihood of confusion from the evidence that there was at least some actual confusion among persons responsible for marketing and processing the orders for the competing check products. 23 H. Summary. Having analyzed the key factors individually, we must now weigh the factors to determine whether there was sufficient evi *980 dence for the jury to find that there is a likelihood of confusion between the marks Memory Stub and Entry Stub as used on the parties’ products. Harland’s mark is not particularly strong and although the two marks are similar in design, they are far from identical. However, the products, the retail outlets and purchasers, and the advertising media used by the parties all are exactly the same. Further, there was at least some evidence from which the jury could infer that Clarke intended to adopt a similar mark in order to benefit from Har-land’s established reputation, and there also was some evidence of actual confusion. Thus, although we would not necessarily have found a likelihood of confusion on these facts ourselves, we hold that there was substantial evidence to support the jury’s finding. Accordingly, we affirm the verdict and judgment on the trademark infringement claim. IV. TRADE DRESS INFRINGEMENT Section 43(a) of the Lanham Act, 15 U.S. C.A. § 1125, states in relevant part: Any person who shall ... use in connection with any goods or services, or any container or containers for goods ... any false description or representation, including words or other symbols tending falsely to describe or represent the same, . .. shall be liable to a civil action by .. . any person who believes that he is or is likely to be damaged by the use of any such false description or representation. This court and the former Fifth Circuit have held that § 43(a) creates a federal cause of action for trade dress infringement. See, e.g., Original Appalachian Artworks, Inc. v. Toy Loft, Inc., 684 F.2d at 830-32; Sun-Fun Products, Inc. v. Suntan Research & Development Inc., 656 F.2d at 192. “Trade dress” involves the total image of a product and may include features such as size, shape, color or color combinations, texture, graphics, or even particular sales techniques. See, e.g., Original Appalachian Artworks, Inc. v. Toy Loft, Inc., 684 F.2d at 831 (“adoption procedures used by [plaintiff] in the sale of its dolls qualify as protectable trade dress”); SK & F Co. v. Premo Pharmaceutical Laboratories, Inc., 481 F.Supp. 1184,1187 (stating that “[t]rade dress is a complex composite of features” including, inter alia, size, color, texture, and graphics, which must “be considered together, not separately”), aff’d, 625 F.2d 1055 (3d Cir.1980); 1 J.T. McCarthy, supra, § 8.1, at 230-31. Most trade dress infringement actions involve the packaging or labeling of goods. See, e.g., Chevron Chemical Co. v. Voluntary Purchasing Groups, Inc., 659 F.2d 695 (5th Cir.1981) (Unit A) (packaging of lawn and garden chemical products), cert, denied, 457 U.S. 1126, 102 S.Ct. 2947, 73 L.Ed.2d 1342 (1982); Sun-Fun Products, Inc. v. Suntan Research & Development Inc., supra, (packaging of sun tan preparations); Perfect Fit Industries, Inc. v. Acme Quilting Co., 618 F.2d 950 (2d Cir.1980) (packaging of mattress pads), cert. denied, — U.S. —, 103 S.Ct. 73, 74 L.Ed.2d 71 (1982). Recently, however, courts have recognized that the design of a product itself may constitute protectable trade dress under § 43(a) of the Lanham Act. See, e.g., Warner Bros., Inc. v. Gay Toys, Inc., 658 F.2d 76 (2d Cir.1981) (distinctive color and symbols on toy car protected under § 43(a)); Harlequin Enterprises Ltd. v. Gulf & Western Corp., 644 F.2d 946 (2d Cir.1981) (distinctive book covers protected against trade dress infringement); Truck Equipment Service Co. v. Fruehauf Corp., 536 F.2d 1210 (8th Cir.) (unique exterior design of twin hopper bottomed grain semi-trailer protected under § 43(a)), cert. denied, 429 U.S. 861, 97 S.Ct. 164, 50 L.Ed.2d 139 (1976). In order to prevail on a claim for trade dress infringement under § 43(a), plaintiff must prove three basic things: “[T]hat the trade dress of the two products is confusingly similar, that the features of the trade dress are primarily non-functional, and that the trade dress has acquired secondary meaning.” Black & Decker Co. v. Ever-ready Appliance Manufacturing Co., 518 F.Supp. 607, 616 (E.D.Mo.1981), aff’d, 684 F.2d 546 (8th Cir.1982); accord, *981 Vuitton Et Fils S.A. v. J. Young Enterprises, Inc., 644 F.2d 769, 772-73 (9th Cir.1981); SK & F Co. v. Premo Pharmaceutical Laboratories, Inc., 625 F.2d 1055, 1065 (3d Cir. 1980); Truck Equipment Service Co. v. Fruehauf Corp., 536 F.2d at 1217-21. We need not discuss the first element of trade dress infringement, likelihood of confusion, in detail. The factors relevant to determining whether there is a likelihood of confusion between the trade dress of two products are “essentially the same” as those relevant to determining whether there is a likelihood of confusion between the parties’ trademarks. Sun-Fun Products, Inc. v. Suntan Research & Development Inc., 656 F.2d at 192; see also Boston Professional Hockey Ass’n v. Dallas Cap & Emblem, Inc., 510 F.2d 1004, 1010 (5th Cir.) (“As a general rule ... the same facts which would support an action for trademark infringement would also support an action for unfair competition.”), cert. denied, 423 U.S. 868, 96 S.Ct. 132, 46 L.Ed.2d 98 (1975). We have already concluded above that these factors supported the jury’s finding that there was a likelihood of confusion between the parties’ marks. Thus, although the scope of the inquiry is somewhat broader under § 43(a), see Sun-Fun Products, Inc. v. Suntan Research & Development Inc., 656 F.2d at 192 (stating that “[t]he touchstone under [§ 43(a)] is not similarity of the registered mark but similarity in the overall trade dress of the products”), we have little difficulty concluding that these factors also support the jury’s conclusion that there was a likelihood of confusion between the trade dress of the Memory Stub and Entry Stub products. 24 We also have little difficulty with the third element of an action for trade dress infringement, secondary meaning. 25 Although Clarke argued at trial that Har- *982 land’s Memory Stub product had not acquired secondary meaning, Clarke has not raised that point on appeal. The crucial question on the trade dress aspect of this case, therefore, is whether Harland has met the second requirement for a trade dress infringement action by demonstrating that any features of the trade dress of its Memory Stub product are “primarily nonfunctional.” 26 If, as Clarke contends, all of the features of Harland’s Memory Stub product are functional, then Clarke was free to copy those features in the absence of any trademark or copyright protection. See Fisher Stoves, Inc. v. All Nighter Stove Works, Inc., 626 F.2d 193 at 195 (1st Cir.1980) (stating that “[functional features which are not the subject of a valid patent or copyright may be imitated with impunity”); Keene Corp. v. Paraflex Industries, Inc., 653 F.2d 822, 824 (3d Cir. 1981) (stating that “[o]ur natural inclination to disapprove of [imitating or copying] must give way to the public policy favoring competition, even by slavish copying, of products not entitled to federal patent protection”); B.H. Bunn Co. v. AAA Replacement Parts Co., 451 F.2d at 1259 (noting that “outright copying is often a civilizing rather than a cannibalizing folkway”). If, however, Clarke copied features of Harland’s trade dress which are primarily nonfunctional, then Clarke is liable for infringement under § 43(a). “The issue of functionality has been consistently treated as a question of fact.” Vuitton Et Fils S.A. v. J. Young Enterprises, Inc., 644 F.2d at 775. Thus, as with the finding of likelihood of confusion discussed in Part III, supra, we must affirm the jury’s verdict if it was based on substantial evidence. 27 Once again the issue is *983 close, but we have concluded that a reasonable jury could find that Clarke copied nonfunctional elements of Harland’s trade dress. Consequently, we must affirm the jury’s verdict finding Clarke liable for trade dress infringement. As the Eighth Circuit once stated, “[t]he line between functionality and nonfunction-ality is not ... brightly drawn in every case.” Truck Equipment Service Co. v. Freuhauf Corp., 536 F.2d at 1218. In this case, we believe that many of the features of Harland’s Memory Stub product — such as the concept and location of the carry-around stub, 28 the lines for recording information on the stub 29 the vertical size of the stub, 30 and perhaps the horizontal size of *984 the stub as well 31 — clearly are functional. A reasonable jury could conclude, however, that some features of the trade dress of Harland’s product — such as the decorative box around the lines for recording information on the stub, 32 and the design and color of the carry-around case 33 — are primarily nonfunctional. Thus, although Clarke was free to copy the functional features of Har-land’s Memory Stub product, we believe the jury could have reasonably concluded that Clarke infringed on the trade dress of Har-land’s product when it produced and marketed its virtually identical product with a similar decorative box on the stub and an almost exact copy of Harland’s carry-around case, thereby creating a likelihood of confusion between the overall trade dress of the competing products. Accordingly, we affirm the verdict against Clark for trade dress infringement. V. THE SCOPE OF INJUNCTIVE RELIEF Clarke’s final argument on appeal is that the permanent injunction entered by the district court does not set forth in sufficient detail the acts which are prohibited. We agree. . Federal Rule of Civil Procedure 65(d) states in relevant part: Every order granting an injunction ... shall be specific in terms [and] shall describe in reasonable detail ... the act or acts sought to be restrained.... The judgment entered by the district court enjoins Clarke from “using trade dress for bank check products which is confusingly similar to the trade dress or overall appearance of plaintiff’s Memory Stub check products or is likely to cause confusion therewith .... ” Record on Appeal, *985 vol. 2, at 1520. We do not believe that this language adequately specifies the acts which Clarke must discontinue in order to avoid infringing on the trade dress of Har-land’s Memory Stub product. As. one commentator has stated: [T]he injunction should clearly let defendant know what he is ordered to do or not to do. A court order should be phrased in terms of objective actions, not legal conclusions. An injunction which merely forbids a defendant from performing “acts of unfair competition,” or from “infringing on plaintiff’s trademarks and trade secrets” adds nothing to what the law already requires. 2 J.T. McCarthy, supra, § 30.8, at 334. Accordingly, we vacate the portion of the district court’s order quoted above and remand to the district court for entry of an order which specifically describes the acts which are prohibited'by the permanent injunction. 34 In all other respects, the judgment of the district court is affirmed. Each party shall bear its own costs on appeal. AFFIRMED IN PART, VACATED AND REMANDED IN PART. OH ACCOUNT TOTAL DEDUCTIONS [[Image here]] Sample Counter Check. (From Defendant’s Exhibit 23). *986 [[Image here]] FIGURE 3 Portion of Clarke’s Entry Stub Product— Check with Intermediate Stub and Permanent Stub attached. (From Plaintiff’s Exhibit 21). [[Image here]] FIGURE 4 Portion of Harland’s Memory Stub Product— Check with Intermediate Stub and Permanent Stub attached. (Composite From Two Exhibits). *987 [[Image here]] FIGURE 5 Portion of Clarke’s Entry Stub Product — Carryaround Case (Plaintiff’s Exhibit 32). *988 [[Image here]] FIGURE 6 Portion of Harland’s Memory Stub Product — Carryaround Case (Plaintiff’s Exhibit 7). Footnotes [1] The check number is imprinted on the check, the intermediate stub, and the permanent stub. [2] Plaintiffs Exhibit 14 (advertising brochure produced by Harland). Harland’s advertising also stated that the Memory Stub was “a great little invention that allows you to write checks at home as you normally would, and to write checks away from home with a minimum of error.” Id. Harland’s use of an intermediate stub was not entirely new, nor was the idea of an end stub to temporarily record information from a transaction. Payroll checks, manufactured by several companies, have long contained an intermediate stub which is designed to be detached, along with the corresponding check, from the permanent check stub. Before cashing the check, the payee detaches the intermediate stub, which contains such information as gross pay, taxes, and net pay. See Figure 1. “Counter checks” issued by banks often have a stub on which information regarding the transaction can be temporarily recorded. See Figure 2. Harland’s innovation was to combine these two concepts, creating an intermediate stub on which information regarding a transaction could be recorded temporarily until the check writer could enter the information on the permanent stub in the desk-style checkbook. [3] Harland marketed the Memory Stub product in three check styles. Two of the styles, the “Classic” and the “Petals,” had artistic background designs, and Harland obtained copyright registrations for those two styles. The third style, the “Ledger,” did not have an artistic background design, and Harland did not register that style. Nevertheless, Harland argues that it copyrighted “its original MEMORY STUB expression ... [as well as] the background designs on two of its three styles.” Brief of Appellee/Cross-Appellant at 3. Clarke argues that Harland’s copyright registrations only protected the background designs. Brief of Appellant/Cross-Appellee at 9-10. [4] William Alfaro, a production and methods manager for Clarke during the period when the new product was being developed, indicated that the initial mock-up of Clarke’s new product actually was one of Harland’s Odyssey Collection products. Record on Appeal, vol. 5, at 219-21. [5] The district court did not submit Harland’s state law claims to the jury, but entered judgment for Harland on those claims after the jury found that Clarke had engaged in trademark and trade dress infringement. [6] Harland does not contend that Clarke has infringed on these artistic background designs. [7] Harland admits that it registered the “Classic” and “Petals” Memory Stub check styles under subsection (k) of section 5 of the old Copyright Act, 17 U.S.C. § 5(k) (1947) (effective for registrations prior to January 1, 1978), which covered “[p]rints and pictorial illustrations including prints or labels used for articles of merchandise.” Nevertheless, Harland argues that “[Registration of a work in one class protects all copyright component parts of that work, even if the other copyrightable components may not fall within that particular class of registration.” Reply Brief of Appel-lee/Cross-Appellant at 2 (emphasis in original). Our conclusion below that the “other components” of the Memory Stub product, besides the background design, were not copyrightable makes it unnecessary for us to consider the validity of Harland’s argument. [8] Harland contends that recent authority has undermined the rule established in Baker v. Selden, supra, and followed by the Copyright Office since that time, citing Norton Printing Co. v. Augustana Hospital, 155 U.S.P.Q. (BNA) 133 (N.D.Ill.1967), and Harcourt, Brace & World, Inc. v. Graphic Controls Corp., 329 F.Supp. 517 (S.D.N.Y.1971). The district court carefully considered and discussed these decisions, distinguishing the former and declining to follow the latter. See John H. Harland Co. v. Clarke Checks Co., 207 U.S.P.Q. (BNA) 664 (N.D.Ga.1980). We agree with the district court’s conclusions and need not repeat that discussion here. Further, we note that the Copyright Office recently reaffirmed the rule of Baker v. Selden, supra, and the regulation quoted in the text, expressly citing the district court’s decision in this case as an example of the correct application of the rule. See Notice of Termination of Inquiry Regarding Blank Forms, 45 Fed.Reg. 63297-300 (1980). [9] In Bonner v. City of Prichard, 661 F.2d 1206 (11th Cir.1981) (en banc), this court adopted as binding precedent all of the decisions of the former Fifth Circuit handed down prior to the close of business on September 30, 1981. Id. at 1209. [10] Clarke’s only objections to the portion of the charge covering trademark infringement are that it was “unduly long and complex” and was “coupled with an incorrect charge on unfair competition.” Brief of Appellant/Cross-Appel-lee at 39. Our review of the record indicates that the district court’s charge regarding the elements of trademark infringement was relatively long, covering 17 pages of transcript, but that the charge correctly stated the law and was not so complex that a reasonable juror would have difficulty understanding it. Further, although we conclude below that the district court could have expanded on its discussion of functionality in the unfair competition portion of the instructions, see note 27 infra, we cannot agree that the charge, as given, incorrectly stated the law. In any event, the district court clearly distinguished the trademark infringement portion of the charge from the portion of the charge dealing with trade dress infringement. Thus, we find no error in the district court’s charge regarding trademark infringement which would warrant reversal. [11] In Boeing Co. v. Shipman, 411 F.2d at 374, the former Fifth Circuit delineated the standards for determining when the evidence is sufficient so that a case must be submitted to the jury. The court stated On motions for directed verdict and for judgment notwithstanding the verdict, the court should consider all of the evidence — not just that evidence which supports the non-mover’s case — but in the light and with all reasonable inferences most favorable to the party opposed to the motion. If the facts and inferences point so strongly and overwhelmingly in favor of one party that the court believes that reasonable men could not arrive at a contrary verdict, granting of the motions is proper. On the other hand, if there is substantial evidence opposed to the motions, that is, evidence of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment might reach different conclusions, the motions should be denied, and the case submitted to the jury. The same standard is applicable on appeal to determine whether the evidence was sufficient to submit the case to the jury. Id. at 367 n. 1; 9 C. Wright & A. Miller, Federal Practice & Procedure: Civil § 2524, at 542 n. 28 (1971). [12] “In order to establish secondary meaning[,] the plaintiff ‘must show that the primary significance of the term in the minds of the consuming public is not the product but the producer.’ ” Vision Center v. Opticks, Inc., 596 F.2d 111, 118 (5th Cir. 1979) (quoting Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 118, 59 S.Ct. 109, 113, 83 L.Ed. 73 (1938)), cert. denied, 444 U.S. 1016, 100 S.Ct. 668, 62 L.Ed.2d 646 (1980). Descriptive marks must have acquired secondary meaning in order to be protected, but arbitrary or fanciful marks and suggestive marks are protected without evidence of secondary meaning. 1 J.T. McCarthy, supra, § 15:1. On appeal, Clarke does not argue that Harland’s Memory Stub mark has not acquired secondary meaning, but simply argues that its Entry Stub mark does not infringe on the Memory Stub mark. [13] Clarke emphasizes that Harland knew in advance that its Memory Stub trademark was rather weak, citing an internal memorandum between two Harland officials. The memorandum began by noting that the Memory Stub mark had not been registered previously, but then stated that Harland’s trademark counsel “feels the name is somewhat weak because it is so suggestive ....” Defendant’s Exhibit 41. (The memorandum then stated, somewhat ironically, that Harland’s primary interest in seeking trademark protection was “to inhibit our competitors from copying the idea rather than take serious legal action.” Id.). Of course, even if Harland’s mark initially was weak, it may have subsequently acquired strength through Harland’s promotional efforts. See Standard Int’l Corp. v. American Sponge & Chamois Co., 394 F.2d 599, 600 (C.C.P.A.1968) *975 (“[A] mark which is initially a weak one may, by reason of subsequent use and promotion, acquire such distinctiveness that it can function as a significant indication of a particular producer as source of the goods with which it is used.”). Harland featured the Memory Stub mark prominently in its advertising and received over 400,000 orders for Memory Stub products during the three years prior to Clarke’s use of the Entry Stub mark. Thus, the fact that Harland’s counsel believed the mark was weak in 1976 does not necessarily suggest that the mark was still weak in 1979. [14] Harland vigorously contends that the two marks carry similar connotations, emphasizing that Clarke even states in its brief that “without the carry-around stub, the check writer would have to rely on their [sic] memory of the transaction unless they [sic] had something to enter it on.” Brief of Appellee/Cross-Appel-lant at . 12 (emphasis in original) (quoting Brief of Appellant/Cross-Appellee at 5). We. doubt that this use of the two words in the same sentence of a legal brief has any significant probative value in determining whether consumers are likely to confuse the two marks. [15] The fact that the names of the parties appear in small print on the checks and the permanent check stubs does not preclude a finding that there is substantial similarity between the design of Harland’s Memory Stub product and the design of Clarke’s Entry Stub product. See note 24 infra. [16] The fact that Clarke copied features of Har-land’s product does not, in itself, entitle Har-land to any relief. “There is nothing wrong with exact copying of an article or a symbol in the public domain. Even if defendant admits that he used plaintiff’s package as a model to copy from in designing his own package, this does not necessarily mean that defendant intentionally was passing off his product as plaintiffs. Such an admission is merely evidence relevant to the ultimate test of likelihood of confusion.” 2 J.T. McCarthy, supra, § 23:35, at 111-12. [17] Harland emphasizes that Clarke “introduced its product with virtually no advertising,” Brief of Appellee/Cross-Appellant at 13, and argues that this fact also suggests that Clarke was attempting to trade on the good will associated with Harland’s Memory Stub mark. Cf. Men-ley & James Laboratories v. Approved Pharmaceutical Corp., 438 F.Supp. 1061, 1067 (N.D.N. Y.1977) (evidence of intent to benefit from good will associated with senior user’s reputation “is especially convincing when the newcomer ... has spent little or nothing at all in promoting its product”). We have not given any weight to this argument because the un-contradicted testimony of Clarke officials indicated that materials advertising the Entry Stub product were withheld due to the present litigation. Thus, Clarke’s “lack of advertising” does not necessarily evidence any intent to deceive or defraud. [18] According to the testimony of James Coin, the president of Clarke, Clarke officials rejected two other suggested marks for the intermediate carry-around stub, including “Reminder Stub,” because their trademark counsel believed that these marks would be likely to cause confusion. Record on Appeal, vol. 6, at 465-69. [19] Frances Lawrence of the Dunwoody, Georgia, Branch of Fulton National Bank testified that she had filled out this order form for a customer and placed it in an envelope for delivery to Clarke checks. (Ms. Lawrence had used one of Harland’s pre-printed order forms, lining out the Harland name, writing in Clarke, and typing the information regarding the “Desk-ette” product and “World of Beauty” background design; this procedure in itself does not suggest any confusion between the Memory Stub and Entry Stub products because Ms. Lawrence’s testimony indicates that she routinely used one check printing company’s pre-printed form to place orders with .another check company.) Ms. Lawrence testified that the card with the words “Cover and Memory Case” was not attached to the order form when she placed it in the envelope. Thus, the card apparently was attached either by one of Ms. Lawrence’s associates at the bank or by a Clarke employee processing the order. [20] Clarke argues that the new accounts representative who completed this form stated in a deposition that there was no confusion in her mind between the Clarke and Harland products and that she simply “wrote down the wrong thing” by mistake. See Record on Appeal, vol. 5, at 236. We do not see the relevance of the distinction Clarke is attempting to draw because the Lanham Act prohibits use of a mark which is “likely ... to cause mistake” as well as a mark which is likely to cause confusion or to deceive. See 15 U.S.C.A. § 1114(1). [21] In both instances, the new accounts representatives clearly intended to place an order with Clarke, not with Harland. At trial, Har-land did produce numerous reorder forms for Memory Stub products which banks had sent to Clarke and which Clarke then had filled with Entry Stub products. This evidence does not necessarily suggest that there was any actual confusion between the products, however, because the record indicates that check printing companies routinely fill orders which have been placed on another company’s reorder form. Assume, for example, that a bank which has been doing business with Harland decides to do business with Clarke instead. Subsequently, many of the bank’s customers will bring in check reorder forms which were prepared by Harland and placed near the back of the package of checks provided by Harland. The bank, however, will send these reorder forms to Clarke, which then will fill the order with a similar product. Most check printing companies, including Harland and Clarke, follow this procedure. Indeed, one of Harland’s plant managers testified that his plant typically receives 50 to 300 orders a day on the order forms of other companies and almost always fills these orders with Harland products. Record on Appeal, vol. 7, at 612-13. Thus, the fact that Clarke received and filled reorders which asked for Harland’s products merely indicates that Clarke was able to produce similar products, not that bank personnel and customers were confused by any similarity between the Entry Stub and Memory Stub marks. [22] One of the problems in determining whether there is a likelihood of confusion in this case is that the parties have not clearly identified the persons who are (or are not) likely to be confused. (This problem also makes it difficult to analyze the degree of care purchasers use when buying the parties’ products. Cf. Sun-Fun Products, Inc. v. Suntan Research & Development Inc., 656 F.2d at 189 (stating that another factor courts should use in assessing the likelihood of confusion is “the degree of care purchasers are likely to exercise when selecting products of the type sold by the parties”)). The rule courts usually apply is that infringement occurs when “there is a likelihood of confusion in the mind[s] of an appreciable number of ‘reasonably prudent’ buyers.” 2 J.T. McCarthy, supra, § 23:27, at 87-88. In this case, however, although the ultimate buyers of the products are checking account customers, banking institutions actually decide whether to carry check products made by Harland, Clarke, or some other check stationery company. Consequently, many, if not most, actual buyers only see (and only can choose from) the products of one check printing company because their bank uses that company’s products exclusively. Moreover, even if a bank carries several companies’ products, customers may not be presented with an opportunity to choose between the products because new accounts personnel at the bank may show them only one company’s catalog. See Record on Appeal, vol. 5, at 232-33 (new accounts representative testified that although her bank carried both Har-land and Clarke checks, she only showed customers the Clarke catalog as she preferred doing business with Clarke). The trial court’s instructions stated that in determining the likelihood of confusion, the jury “should consider consumers, including purchasers or users of the parties’ products, and ... should also consider persons who assist consumers in selecting and ordering the goods they wish to purchase.” Record on Appeal, vol. 7, at 868. Neither party has objected to this charge on appeal. In the absence of more detailed information regarding the marketing of bank check products, we assume that this portion of the instructions was correct. [23] Both parties also presented “expert” witnesses on the likelihood of confusion issue. Harland’s expert, Fred Corley, president of the Midland City Bank in Midland, Alabama, testified that he believed that there was a likelihood of confusion between the Memory Stub and Entry Stub products, although he was unaware of any actual confusion among his bank’s customers or employees. Clarke’s expert, Abraham Beaugarad, a trademark lawyer from Washington, D.C., testified that there was no likelihood of confusion between the two products, but admitted that he had no background in the check printing business and had not conducted any survey regarding the likelihood of confusion among consumers. We do not believe that the testimony of either expert is entitled to much weight in analyzing whether the evidence was sufficient to support the jury’s verdict. [24] Clarke points out that the Clarke name appears in the lower left comer of each check and the lower left corner of each permanent stub, as well as on notices and packaging materials which are sent to customers and on a calendar which is placed in the desk-style checkbook. Clarke contends that these facts require reversal of the judgment for trade dress infringement, citing Bose Corp. v. Linear Design Labs, Inc., 467 F.2d 304 (2d Cir. 1972), for the proposition that “when a competitor marks its products with its name no cause of action ... under § 43(a) can lie.” Brief of Appellant/Cross-Ap-pellee at 15. We disagree. The presence of a manufacturer’s name on a product is just one of many factors relevant to determining whether there is a likelihood of confusion. See, e.g., Truck Equipment Service Co. v. Fruehauf Corp., 536 F.2d at 1214 n. 2, 1220-21 (fact that defendant labeled trailer as its own product did not preclude action under § 43(a) or make lower court’s finding of likelihood of confusion clearly erroneous); cf. 2 J.T. McCarthy, supra, § 23:15(H), at 61 (“The fact that defendant places its own name on a label together with the symbol alleged to be an infringement does not dictate that there can be no possible confusion.”). In fact, the subsequent history of the case cited by Clarke demonstrates that placement of the manufacturer’s name on a product does not preclude a finding of infringement. See Bose Corp. v. Linear Design Labs, Inc., 194 U.S.P.Q. (BNA) 339 (S.D.N.Y.1977) (noting that a consent judgment of trademark infringement was entered on remand). In this case, Clarke’s name appears in very small print on the checks and permanent stubs. Cf. SK & F Co. v. Premo Pharmaceutical Laboratories, Inc., 625 F.2d at 1061 (affirming district court’s finding that logos on capsules “were so small that they would be ineffective to prevent confusion as to source or passing off’). In addition, Clarke’s name appears on a variety of other materials, much of which is not received by the ultimate consumer until after an order is placed. Thus, although such evidence is relevant to the likelihood of confusion issue, we do not believe that it precludes a finding of trade dress infringement. [25] Several courts, including the former Fifth Circuit in a decision which is not binding on us, have held that the plaintiff in a trade dress infringement action need not prove secondary meaning when the product’s trade dress is inherently distinctive. See, e.g., Chevron Chemical Co. v. Voluntary Purchasing Groups, Inc., 659 F.2d at 702 (stating 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, there is no reason to require a plaintiff to show consumer connotations associated with such arbitrarily selected features.”); Perfect Fit Industries, Inc. v. Acme Quilting Co., 618 F.2d at 952-54 (concluding that, under New York law, a plaintiff need not prove secondary meaning to obtain relief for trade dress infringement when the trade dress, in itself, is distinctive and memorable). Although we believe this approach has merit, we need not consider it to resolve the case before us. [26] We recognize that there is some confusion in the courts regarding whether the plaintiff has the burden of proving nonfunctionality as part of the prima facie case or whether the defendant has the burden of proving functionality as a defense. Compare, e.g., Vibrant Sales, Inc. v. New Body Boutique, Inc., 652 F.2d 299 (2d Cir.1981) (placing burden on plaintiff), cert. denied, 455 U.S. 909, 102 S.Ct. 1257, 71 L.Ed.2d 448 (1982), with Ives Laboratories, Inc. v. Darby Drug Co., 601 F.2d 631, 643 (2d Cir.1979) (stating, in affirming the denial of a preliminary injunction, that “[t]he case for functionality ... depends on the evidence proffered by defendants that copying [certain colors] served a number of utilitarian purposes”). Cf. Fisher Stoves, Inc. v. All Nighter Stove Works, Inc., 626 F.2d 193, 195-96 (1st Cir.1980) (expressing some doubt about where the burden should be placed). In this case, the district court instructed the jury that Harland had the burden of proving nonfunctionality, and Harland has not objected to that instruction on appeal. Thus, we need not decide whether the instruction correctly stated the law. Cf. Note, The Problem of Functional Features: Trade Dress Infringement Under § 43(a) of the Lanham Act, 82 Colum.L.Rev. 77, 87 n. 78 (1982) [hereinafter cited as The Problem of Functional Features ] (arguing that the defendant should have the burden of proving functionality). [27] The jury’s verdict need not be affirmed, of course, if the jury was not properly instructed, and Clarke makes a strong argument that the jury was inadequately instructed on the issue of functionality. Our review of the record indicates that the trial court clearly instructed the jury that it must find that Clarke had copied nonfunctional features of Harland’s product before it could find trade dress infringement. The instructions stated in relevant part: I charge you that the fact that the defendant utilizes functional features on its Entry Stub check product that are similar or even identical in some respects to functional features on plaintiff’s Odyssey Collection check product does not and cannot constitute unfair competition. Now, ladies and gentlemen of the jury, in order to establish its claim of [trade dress infringement], plaintiff must prove by a preponderance of the evidence that, (1), the design aspects or trade dress of Harland’s Memory Stub product are primarily nonfunctional. ... Thus, the first issue that you must decide is whether or not the trade dress of Harland’s Memory Stub product is primarily nonfunctional. In evaluating this issue, you must keep in mind that even though an object may serve some practical purposes, its design is protectable as trade dress if you determine the design is primarily nonfunctional. Now, ladies and gentlemen, after you have determined which, if any, aspects of the trade dress of Harland’s Memory Stub products are primarily nonfunctional, you must then consider whether the trade dress of Harland’s products [has acquired secondary meaning]. Record on Appeal, vol. 7, at 871-72. The judge’s instructions correctly stated the current law regarding the role of functionality in trade dress infringement actions. See, e.g., Vuitton *983 Et Fils S.A. v. J. Young Enterprises, Inc., 644 F.2d at 772-75; Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 203-05 (2d Cir.1979). Clarke contends, however, that the trial judge erred by failing to give a proffered instruction which included a detailed definition of the term functionality. We admit that the trial judge could have expanded on the instructions regarding the doctrine of functionality, but we reject Clarke’s contention that the judge’s failure to do so constitutes reversible error. As noted above, the instructions actually given correctly stated the law. Further, our review of the entire record, including the evidence presented during the trial, the arguments of counsel for both parties, and the instructions given, indicates that the jury was adequately informed about the role of functionality in a trade dress infringement action. Cf. First Virginia Bankshares v. Benson, 559 F.2d 1307, 1316 (5th Cir.1977) (stating that the standard test for reviewing the trial court’s instructions is to “consider the challenged instruction as part of the entire charge, in view of the allegations of the complaint, the evidence presented, and the arguments of counsel, to determine whether the jury was misled and whether the jury understood the issues”), cert. denied, 435 U.S. 952, 98 S.Ct. 1580, 55 L.Ed.2d 802 (1978); see also Shelak v. White Motor Co., 581 F.2d 1155, 1160-61 (5th Cir.1978). Moreover, Clarke’s proffered instruction contained an incorrect definition of functionality. The proffered instruction stated: “When a feature of goods or of its wrappers or containers appeals to the consumer and affects his or her choice, that feature is functional.” Record on Appeal, vol. 2, at 1451 (Defendant’s Request to Charge No. 38) (emphasis in original). We believe that this definition is overly broad because many nonfunctional, arbitrary features of a product may appeal to the consumer and affect his or her choice. Our conclusion is supported by the Ninth Circuit’s recent rejection of a similar definition of functionality in Vuitton Et Fils S.A. v. J. Young Enterprises, Inc., supra. The district court had found that a repeated fabric design on plaintiffs luggage products was functional because it constituted “the primary decoration of those goods” and was “a factor in their consumer appeal and saleability.” 644 F.2d at 773. The Ninth Circuit reversed, stating in relevant part: “We disagree with the district court insofar as it found that any feature of a product which contributes to the consumer appeal and saleability of the product is, as a matter of law, a functional element of that product.” Id. The Ninth Circuit also rejected the argument that a design is functional if it is related to the reasons consumers purchase the product, noting that even trademarks would be functional under such a definition because they help to sell goods by identifying the manufacturer. Id. at 774. See also The Problem of Functional Features, supra note 26, at 88 (criticizing a similar definition of functionality, any product element which is “an important ingredient in the commercial success of the product,” as overly broad); cf. Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 850 n. 10, 102 S.Ct. 2182, 2186 n. 10, 72 L.Ed.2d 606, 613 n. 10 (1982) (“In 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.”). [28] Harland’s own marketing materials indicate that the concept of an intermediate carry-around stub, designed to solve the portability problem associated with desk checkbooks, is functional. See note 2 and accompanying text supra. Harland cannot appropriate this functional innovation under the guise of trade dress. Further, although Harland has argued that Clarke could have placed the carry-around stub in a different location, such as at the top or the bottom of the check, we believe the evidence demonstrates that the placement of the carry-around stub between the check and the permanent stub is functional. [29] The lines for recording the date, the amount, the payee and the purpose of the check correspond to the lines for recording information on traditional check stubs and are functional. [30] The vertical size of the stub is functional because it must correspond to the vertical size of the check and the permanent stub. Further, although it may be technically feasible for Clarke to vary the vertical size of all three components, i.e., the check, the permanent stub and the carry-around stub, the record suggests that the size Clarke selected is standard throughout the industry. For example, the check and permanent stub in Clarke’s Entry Stub product are the same size as the check *984 and permanent stub in Clarke’s traditional three-check-to-a-page checkbook, which was on the market long before Harland’s Memory Stub product. Thus, even if Clarke could have varied the size of its check and permanent stub without affecting the utility of the product or the production costs, Harland could not claim that the size of these components was protecta-ble trade dress because no reasonable jury could conclude that this particular check size had acquired secondary meaning. [31] Harland vigorously argues that the horizontal size of its carry-around stub is nonfunctional, and we agree that Clarke probably could have made its carry-around stub wider without affecting its utility or the production costs. There are functional limits on the horizontal size of the carry-around stub, however, and those limits would soon be reached if each check printing company was required to widen its stub sufficiently to distinguish it from stubs which already had been successfully marketed by its predecessors. Cf. Morrissey v. Procter & Gamble Co., 379 F.2d 675, 678 (1st Cir.1967) (“When the uncopyrightable subject matter is very narrow, so that ‘the topic necessarily requires’ ... if not only one form of expression,' at best a limited number, to permit copyrighting would mean that a party or parties, by copyrighting a mere handful of forms, could exhaust all possibilities of future use of the substance.”) (citations omitted; quoting Sampson & Murdock Co. v. Seaver-Radford Co., 140 F. 539, 541 (1st Cir.1905)). Moreover, the size and shape of the carry-around stub may be functional because of the nature of the check printing industry. As discussed in note 22, supra, check companies routinely fill reorders from consumers who initially had obtained their check products from a competitor. It would be very inefficient if each such reorder for desk-style checkbook checks with carry-around stubs had to include a new carry-around case, and possibly even a new desk-style checkbook, in order to accommodate the different-sized carry-around stubs. [32] There is no evidence in the record indicating that the decorative box which appears on both parties’ products serves any functional purpose. [33] Clarke’s carry-around case appears to be an exact copy of Harland’s case except that it is made from a slightly darker brown vinyl. Admittedly, the carry-around case serves a function, but the particular design of the case is not functional. Cf. Dallas Cowboys Cheerleaders v. Pussycat Cinema Ltd., 604 F.2d at 203 (“It is well established that, if the design of an item is nonfunctional ..., the design may become a trademark even if the item itself is functional.”). In fact, Clarke has developed and marketed a differently designed carry-around case since this litigation began. Further, Clarke easily could have chosen a different color for its carry-around case rather than the dark brown which makes Clarke’s case almost indistinguishable from Harland’s dark brown case. [34] In fashioning its order, the district court will of course be'guided by this opinion, including our holding in Part IV supra that certain features of the trade dress of Harland’s product are functional and, therefore, are not protecta-ble under the Lanham Act.