TJOFLAT, Circuit Judge:
Skating Clubs of Georgia, Inc. appeals from a district court order enjoining it from using "Lollipops" as a name for one of its roller skating rinks because that name is too easily confused by consumers with the service mark
In November 1979, Jellibeans, Inc. opened a roller rink in Atlanta, Georgia, named "Jellibeans." It advertised Jellibeans extensively, spending over $50,000 in November and December alone, and over $100,000 during the first year of the rink's operation. Jellibeans, Inc. spent most of this money on radio advertisements, and the rest on newspaper advertisements and local promotional activities. Jellibeans, Inc.'s extensive advertising, and its use of a candy name for its rink, were unique among Atlanta roller rinks. The rink was a success from the outset.
Skating Clubs of Georgia, Inc., operates six roller rinks in the Atlanta area. Skating Clubs established its first rink in 1959, another in 1971, two in 1973, and a fifth rink in 1975. It called each of these rinks
Several patients and a business acquaintance of Dr. Ronald Feinman, a principal shareholder of Jellibeans, Inc., saw the Lollipops sign.
On September 26, 1980, Jellibeans, Inc. brought a four-count complaint in the district court alleging that Skating Clubs' use of the name Lollipops: (1) constituted a false description or representation of Skating Clubs' services, in violation of the Lanham Act, 15 U.S.C. § 1125(a) (1976);
At trial, Jellibeans, Inc. called two patients and a business acquaintance of Dr. Feinman to testify. They stated that they saw Lollipops' sign, believed that Jellibeans, Inc. was opening a new rink because of the similarity of the names, and called Dr. Feinman to inquire whether he was opening a new rink. Jellibeans, Inc. also produced a survey that showed 96% of those interviewed had heard of Jellibeans and 48% of them believed that the Lollipops rink was related to it. Skating Clubs objected to the introduction of this survey in evidence, contending that it was so flawed that it lacked probative value.
In its defense, Skating Clubs presented the testimony of owner, Albert Couey, and his nephew Jerry Taylor, who brokered the sale of the real estate where Lollipops was being built and who served as Skating Clubs' insurance agent. Both Couey and Taylor testified that Taylor suggested the name Lollipops. Both denied that they had considered the similarity of the names Lollipops and Jellibeans. They admitted, however, that they had been aware of Jellibeans. Couey testified that the original sign announcing the construction of Lollipops stood only two days before it was burned by a fire of unknown origin. He testified that Skating Clubs replaced it with an identical sign, except for the word "Disco" which was omitted from the name of the rink. Jellibeans, Inc. materially impeached this testimony, however, by introducing pictures of the two signs that showed the substantial differences between the two, particularly with regard to the ownership of the rink. Finally, Couey testified that he promoted, and would in the future promote, the Lollipops rink only by circulating brochures to local schools and PTA's, giving luncheons and teas for community groups, and newspaper advertisements.
On May 15, 1981, the district court rendered its decision. It dismissed the state law trademark infringement count because Jellibeans, Inc. had not registered its service mark in Georgia.
Skating Clubs appeals. It contends that the district court was clearly erroneous in finding that rollerskaters are likely to confuse Lollipops with Jellibeans.
We note subject matter jurisdiction in this case. Jurisdiction of Jellibeans, Inc.'s Lanham Act claim is conferred by the Lanham Act itself, 15 U.S.C. § 1121.
We are, accordingly, required to determine in this case whether: (1) the district court's finding of fact as to the amount of interstate contact can be accepted as not clearly erroneous; and (2) whether as a matter of law the interstate contacts are sufficient to satisfy the jurisdictional requirements of the Lanham Act, 15 U.S.C. §§ 1121, 1125. The district court found that Jellibeans drew patrons from an area that was within 60 miles of Atlanta, and included parts of Alabama, as well as receiving some out-of-state convention business. In addition, the trial court found that Jellibeans received some free publicity in national magazines. None of the evidence
The district court held that the Georgia deceptive trade practices and unfair competition counts involved the same dispositive question as the federal Lanham Act count. This is a question of state law that the parties do not challenge on appeal. Therefore, we treat this holding as correct
The law of trademark and service mark infringement is part of "the rather swampy area of unfair competition." John H. Harland Co. v. Clarke Checks, Inc., 711 F.2d 966 (11th Cir.1983) (quoting B.H. Bunn Co. v. AAA Replacement Parts Co., 451 F.2d 1254, 1258 (5th Cir.1971) (Goldberg, J.)). A service mark is a form of property and has property rights associated with it. The definition of the property right, where it begins and ends, is the correlative of the question of whether a given act infringes on the property right. Determining whether an infringement has taken place is but the obverse of determining whether the service mark owner's property right extends into a given area. The property right entailed in a service mark has an amorphous, multi-dimensional character that is subject to change with circumstances. The factual setting surrounding the mark, including how it is created, where it is located, and how it is putatively infringed is crucial to determining whether such infringement has taken place. There is no simple, or for that matter complicated, test by which the question of infringement is resolved.
The test for a service mark infringement under 15 U.S.C. § 1125 has often been expressed as whether or not the offending mark is "likely to cause confusion."
The standard of review applicable to the district court's finding of a likelihood of confusion is dependent upon whether this is a question of law, or of fact, or a hybrid embodying both. This circuit has held that it is a question of fact, Harland, 711 F.2d at 972-973, and can be set aside only if clearly
The "likely to confuse" test is not as unidimensional as its appellation would imply. While there are no subparts to this test in the sense of necessary and sufficient conditions there are a number of factors which may bear on the question of likelihood of confusion. In Safeway, this court enunciated a number of such factors: (1) How distinctive is the plaintiff's mark? Do consumers strongly identify plaintiff's service mark with his services? This is important because the more distinctive the plaintiff's mark, the stronger it is considered, and the more protection it is accorded from confusingly similar marks; (2) How similar are the designs of the plaintiff's and defendant's marks? Obviously, the greater the similarity, the greater the likelihood that purchasers will confuse the plaintiff's and defendant's services. This principle also applies to subsidiary facts 3, 4, and 5; (3) How similar are the services the plaintiff's and defendant's marks represent; (4) How similar are the plaintiff's and defendant's retail outlets and their customers; (5) How similar is the advertising the plaintiff and defendant use; (6) Whether the defendant intended that purchasers would confuse his service mark with the plaintiff's mark; and (7) Whether people were actually confused by the similarity of the marks. See Safeway, 675 F.2d at 1164 (and cases cited therein). The court evaluates and weighs these subsidiary findings to determine, as a matter of fact, whether consumers are likely to confuse the defendant's services with the plaintiff's services.
Since it is undisputed that the district court identified this analysis as the correct standard for deciding Jellibeans' Lanham Act claim, our function on review is merely to determine whether the district court properly applied this factual analysis. More specifically, we must determine whether the district court was clearly erroneous in finding the subsidiary facts listed above, and in finding the ultimate fact — whether consumers are likely to confuse Lollipops with Jellibeans. See id. at 1164 n. 5; see Exxon Corp. v. Texas Motor Exchange, 628 F.2d 500, 506 (5th Cir.1980); Armstrong Cork Co. v. World Carpets, Inc., 597 F.2d 496, 501-02 (5th Cir.1979), cert. denied, 444 U.S. 932, 100 S.Ct. 277, 62 L.Ed.2d 190 (1979). We can reverse the district court's decision only if after reviewing all the evidence, we are "left with the definite and firm conviction that a mistake has been committed," Safeway, 675 at 1164, quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 541-42, 92 L.Ed. 746 (1948), by the district court in finding the ultimate fact issue. We can reach this decision in two ways. First, we can determine that the court clearly erred in finding one or more of the subsidiary facts, and that this mistake caused it clearly to err in its decision on the ultimate fact issue. In addition, we can determine that the district court did not
With regard to how distinctive a mark Jellibeans is, the court noted that Jellibeans is an arbitrary mark,
We have no difficulty upholding this finding. Jellibeans clearly is an arbitrary mark as applied to roller rinks. Survey evidence introduced by Jellibeans, Inc. showed that 96% of those interviewed had heard of Jellibeans, indicating strong recognition of the mark. Finally, there is ample evidence in the record that Jellibeans is the only roller rink in the United States using a candy name. In light of this evidence, we hold that the district court's finding that Jellibeans is a very distinctive mark entitled to broad protection is not clearly erroneous.
The second subsidiary fact that the district court examined is the similarity between
After examining the names visually and aurally, we are not left with the definite and firm conviction that the district court made a mistake in finding them similar. Moreover, we cannot say that the district court erred in finding that the two names evoke the same general impression.
We also uphold the district court's findings on the third and fourth subsidiary facts. The district court was not clearly erroneous in finding that both Jellibeans and Lollipops are skating rinks with a clientele primarily composed of teenagers and young adults. Even if, as Skating Clubs alleges, the rinks differ physically and attract patrons of different ages, we find sufficient overlap in their services and clientele to uphold the district court's findings. See Safeway Stores, 675 F.2d at 1166.
The fifth fact the district court examined is the similarity of the advertising used by Jellibeans, Inc. and Skating Clubs. The court found that Jellibeans, Inc. primarily uses radio advertising, supplemented with school and charitable promotions. The court found that Skating Clubs uses, or will use, similar forms of advertising. More specifically, it accepted Skating Clubs' claim that it advertised, and would advertise,
We cannot say that the district court was clearly erroneous in making this finding. The evidence of Jellibeans, Inc.'s advertising practices, which include school and charitable promotions, is uncontradicted. Skating Clubs' use, and intent to use these forms of advertising is also uncontradicted. The only evidence that Lollipops' advertising practices would differ from Jellibeans' is Albert Couey's testimony that Lollipops would not use radio advertising, which the district court rejected. Determining the credibility of this testimony is a task solely within the province of the trial court, which had the opportunity to hear the witness and view his demeanor while testifying. We do not think the trial court's finding on the credibility or lack thereof to be accorded this testimony is clearly erroneous. See United States v. Reddoch, 467 F.2d 897, 898 (5th Cir.1972). Given Couey's other testimony about the content of the two Lollipops signs, which Jellibeans, Inc. materially impeached, we are not left with the definite and firm conviction that the trial court erred in disbelieving Mr. Couey's testimony about Lollipops' advertising. Given this conclusion, we also cannot find that the trial court erred in finding that the parties advertise similarly.
The district court next examined Skating Clubs' intent, and found that Skating Clubs selected its service mark, Lollipops, with the intent of confusing consumers with Jellibeans, Inc.'s successful mark, Jellibeans. The court based this conclusion solely on circumstantial evidence, rejecting Couey's and Taylor's denials that they had selected the name Lollipops with no intent of benefitting from its similarity to Jellibeans. Intent, as the district court noted, must often be proved by circumstantial evidence. Certainly the record in this case — which includes testimony that Skating Clubs' other rinks were called "skating clubs" and named after their geographic location, and that both Couey and Taylor knew of Jellibeans and its success when naming Lollipops — provides sufficient evidence to support a finding that Skating Clubs intended to confuse rollerskaters when it adopted the name Lollipops. Moreover, the court also had more than adequate evidence to disbelieve Couey's and Taylor's denials of improper intent. As stated supra, Jellibeans, Inc. materially impeached Couey's testimony by showing the substantial differences in the Lollipops' signs that Couey described as identical. Jellibeans, Inc. also impeached Taylor's testimony by showing his business and familial relationship with Couey. Given this evidence, we cannot hold that the district court clearly erred in finding that Skating Clubs intended to capitalize on Jellibeans' reputation when it named its new rink Lollipops.
The final fact the district court examined is whether people were actually confused by the similarity of the marks. The court found that such confusion between the marks did exist. It based this finding on two types of evidence: first, the testimony of two patients and one business acquaintance of Dr. Feinman, one of Jellibeans, Inc.'s major stockholders, that they saw the Lollipops sign, thought that Feinman's corporation was opening a new rink, and inquired of or congratulated him on the new rink; second, a survey introduced by Jellibeans, Inc. that showed 48% of those interviewed believed that Lollipops was connected to Jellibeans. The court placed minimal weight on the survey evidence, finding it "seriously flawed," but found that it, taken together with the witnesses'
Skating Clubs cites Sun Banks v. Sun Federal Savings & Loan, 651 F.2d 311 (5th Cir.1981), and Amstar, 615 F.2d 252, to support its argument that three reports of actual confusion are insufficient to support a finding of actual confusion. In Sun Banks, the court held that nineteen reports of actual confusion over a three-year period were insufficient to establish a finding of actual confusion. Similarly, in Amstar the court held that three instances over a fifteen-year period were insufficient proof of actual confusion. Skating Clubs argues that if the evidence offered in those cases was insufficient, the evidence in this case is also insufficient. We disagree.
Skating Clubs' argument presumes that the reviewing court should review the evidence on some sort of absolute scale, rather than examining the evidence in light of the circumstances of the particular case. It is this latter mode of analysis, however, that Sun Banks and Amstar prescribe. In both of those cases, the court specifically qualified its insufficiency finding to the facts of the case. In Sun Banks, for instance, the court stated that the nineteen incidents were insignificant given the extent of the parties' advertising and the number of transactions they handled during the three-year period. In Amstar, the court likewise found that the three instances of actual confusion over fifteen years were insufficient given the parties' extensive advertising. Thus, Sun Banks and Amstar emphasized that in determining the sufficiency of the evidence to prove actual confusion, a court should examine the totality of the circumstances to determine how likely instances of actual confusion would be reported. This examination may include consideration of the time period in question and how extensively the product is advertised or made known to the public, see Sun Banks, 651 F.2d 311; Amstar, 615 F.2d 252, as well as the type of confusion that exists and who suffers the confusion. See Armstrong Cork Co., 597 F.2d 496; Holiday Inns, Inc. v. Holiday Out, 481 F.2d 445 (5th Cir.1973).
Considering the testimony of the three witnesses in this manner, we have no difficulty according it probative value. When the witnesses saw the Lollipops sign, it had been up only a short time and was Skating Clubs' only advertisement. Therefore, few reports of actual confusion could be expected. Cf. 3 R. Callman, The Law of Unfair Competition, Trademarks and Monopolies § 80.6 (3rd ed. 1969) ("when equitable relief [from the defendant's use of a confusingly similar service mark] is sought with due promptitude, the use of defendant's mark will have been of such duration that, even if actual confusion has occurred, proof thereof is virtually unattainable."). Moreover, the reports of confusion that do exist appear genuine, given that the three witnesses independently and on their own initiative contacted Dr. Feinman to inquire about the new rink. Thus, this testimony alone might be sufficient to support a finding of actual confusion. We need not decide this question, though, because the record also contains survey evidence on this issue.
Skating Clubs argues, however, that the district court should not have admitted this survey evidence because of its alleged technical deficiencies: (1) poor sampling; (2) inexperienced interviewers; (3) poorly designed questions; and (4) other errors in execution. These alleged technical deficiencies affect the survey's weight, however, and not its admissibility. See Exxon Corp. v. Texas Motor Exchange, 628 F.2d 500, 507 (5th Cir.1980); Amstar, 615 F.2d at 264; Holiday Inns, Inc., 481 F.2d at 447 ("the district court properly admitted the survey evidence in this case, leaving the format of the questions and the manner of conducting the survey for consideration as to the weight of the evidence."); see also C.A. May Marine Supply Co. v. Brunswick Corp., 649 F.2d 1049, 1055 n. 10 (5th Cir.1981) (court excluded survey evidence because
The question before us then is one of weight — whether this survey evidence combined with the testimony of the three witnesses sufficiently supports a finding of actual confusion. In answering this question, we note that the quantum of evidence needed to show actual confusion is relatively small. Safeway Stores, 675 F.2d at 1167; Roto Rooter Corp. v. O'Neal, 513 F.2d 44, 47 (5th Cir.1975); World Carpets, Inc. v. Dick Littrell's New World Carpets, 438 F.2d 482, 489 (5th Cir.1971). We believe that the evidence in this case satisfies this minimal requirement. As stated supra, the testimony of the witnesses is probative of actual confusion. Notwithstanding its technical deficiencies, the survey evidence is also probative.
Having found the seven subsidiary findings not clearly erroneous, we now must determine whether the district court's evaluation of these findings and its decision on the ultimate fact question — whether roller skaters are likely to confuse Lollipops with Jellibeans — is clearly erroneous. We hold that it is not. Jellibeans, Inc.'s service mark is distinctive and is entitled to broad protection; there is a great similarity between Lollipops' and Jellibeans' designs, services, retail outlets and purchasers, and advertising; and Skating Clubs intended to use these similarities to confuse others, which it did. Weighing these factors together, we conclude that there is a strong likelihood that roller skaters will confuse Lollipops with Jellibeans. We therefore affirm the district court's decision that Skating Clubs violated section 1125 of the Lanham Act. Accordingly, we also affirm the district court's decision that Skating Clubs violated the Georgia Uniform Deceptive Trade Practices Act and engaged in unfair competition under Georgia common law.
Jellibeans, Inc. cross appeals the district court's denial of its claim for attorney fees. Having found that Skating Clubs intentionally adopted a mark confusingly similar to its own mark, Jellibeans, Inc. argues that the district court should have awarded it counsel fees. We disagree.
Under the Lanham Act, a district court may award attorney fees to the prevailing party "in exceptional cases." 15 U.S.C. § 1117 (1976). "Exceptional cases" are those that involve "acts which the courts have characterized as malicious, fraudulent, deliberate, and willful." S.Rep. No. 93-1400, 93rd Cong. 2d Sess. (1974), reprinted in 1974 U.S.Code Cong. & Admin.News 7132, 7136. See Safeway Stores, 675 F.2d at 1169; Vermont Castings, Inc. v. Evans Products Co., No. 79-265 (D.Vt. Dec. 23, 1981); Clairol, Inc. v. Save-Way Industries, Inc., 210 U.S.P.Q. 459 (S.D.Fla.1980); O'Brien International, Inc. v. Mitch, 209 U.S.P.Q. 212 (N.D.Cal.1980); Salton, Inc. v. Cornwall Corp., 477 F.Supp. 975, 992 (D.N.J.1979); RCA Records v. Kory Records, Inc., 197 U.S.P.Q. 908 (E.D.N.Y.1978). Even if the court determines the case to be exceptional, the decision to award attorney fees is still within the court's discretion. See
The district court denied Jellibeans, Inc. attorney fees because it found that Skating Clubs' "intent amounted more to imitation of a successful mark in order to engender similar good will than to actual conversion of the mark." This statement is a finding by the district court that Skating Clubs' infringement of Jellibeans, Inc.'s mark was not malicious, fraudulent, deliberate, or willful, but rather an honest attempt lawfully to copy a successful mark, which unfortunately for Skating Clubs proved unlawful. We cannot say that the district court erred in this finding. Moreover, even if we could hold that the court erred in this finding, we certainly could not hold that the district court abused its discretion in denying the fees in this case. We, therefore, affirm the district court's denial of attorney fees.
Jurisdiction of Jellibeans' claim existed in the district court for the unfair competition count under 28 U.S.C. § 1338(b), and for the two Georgia statutory counts under the doctrine of pendent jurisdiction as diversity is lacking (both parties being Georgia corporations).
Tisch Hotels, Inc. v. Americana Inn, Inc., 350 F.2d 609, 611 n. 2 (7th Cir.1965).
The obvious reason for giving less protection to generic terms in a service mark than to arbitrary terms is that to do otherwise would place an unreasonable burden on third parties. To exclude other providers of a good or service from the right to use generic terms once they have been appropriated by a provider of the good, would quickly result in a situation in which businesses were struggling to convey information about their business in their mark but without using any words that described their business, because those words would already be protected in the service mark of another firm. In such a world, there would be almost no possibility of confusing one firm's service mark with another. Unfortunately the public would also be receiving very little information about the nature of a business' service from its mark.
Therefore it seems that the true legal standard is not simply the "likelihood of confusion" but rather the "likelihood of unreasonable confusion." "Reasonable" confusion is generated by the legitimate efforts of a business to convey vital information to the public about the basic nature of one's business, and to invite comparison with one's competitors. Thus implicit in the legal standard is a balancing between the confusion created between the putatively offending service mark and the burden placed on its owner to find a mark which can convey information to the public about the nature of his service.
The technical deficiencies alleged by Skating Clubs merely reduced the weight of this evidence. For instance, the fact that the survey interviews occurred only at six shopping centers and involved only roller skaters aged 12 to 24 may not provide a truly representative survey sample. It may be that the rollerskating universe is not limited to this age group and does not frequent these shopping centers. Even if this is so, the survey does provide some indication of the substantial confusion experienced by this segment of the population — which Jellibeans, Inc., claims primarily composes its clientele and which Skating Clubs has not contradicted. Therefore, notwithstanding this alleged deficiency, the survey is still probative.
Similarly, the fact that the three interviewers may have been inexperienced does not wholly destroy the probativeness of the survey. These interviewers were college seniors majoring in marketing with top marks in their survey course. Moreover, they were given 90 minutes of instruction on how to conduct the survey. Finally, the interview itself merely required the interviewer to read the questionnaire to the interviewee and properly record his answer. This task does not appear beyond the competence of the "inexperienced" interviewers. Skating Clubs also argues that because four of the 214 questionnaires unexplainedly have an unknown person's initials on them, there must have been a fourth untrained interviewer. Even if this is true, it does not trouble us. The remaining 210 properly recorded questionnaires still remain probative.
The allegation that the survey questions may have been poorly designed also does not totally destroy the probativeness of the survey. Although the survey could have included pictures of Lollipops' and Jellibeans' logos, we believe that a comparison of the names alone deserves some weight. Skating Clubs disagrees, arguing that "word association" surveys are entitled to no weight. Skating Clubs cites Amstar, 615 F.2d 252, and Holiday Inns, Inc., 481 F.2d 445, to support this argument. Skating Clubs, however, reads these cases too broadly.
In Holiday Inns, Inc., the Holiday Inns hotel company claimed that the name Holiday Out, when used by a campground company, was too easily confused with its service mark. To prove actual confusion, Holiday Inns introduced a survey that recorded people's responses when shown a placard bearing the words "Holiday Out" and asked what other company it brought to mind. The Holiday Inns, Inc. court held that this survey merited little weight "because the format failed to account for the number of responses attributable to the word `Holiday' as distinguished from the service mark Holiday Out." 481 F.2d at 448. Thus, the Holiday Inns, Inc. court did not hold that all word association surveys are without weight. Rather, it held that if a service mark contains a word that independently may be associated with other names, the survey should distinguish between the confusion caused by the word, and the confusion caused by the mark that contains the word. If this distinction is not made, the survey is accorded little weight.
In Amstar, that is exactly what happened. There the Amstar Corporation claimed that the "Domino's Pizza" was too easily confused with its service mark, "Domino Sugar." In its survey, Amstar Corporation showed persons the Domino's Pizza mark and asked if it brought anything else to mind; it did not try to account for the number of responses attributable to the word "Domino's." For that reason, inter alia, the Amstar court gave little weight to the survey.
Thus, the mere fact that a survey involves a word association test does not mean that it deserves little weight. As long as no words in the mark blur the results of the test, the survey merits some weight. Since the marks involved in this case contain no such words, the survey is entitled to some weight.
Certain alleged errors in the execution of the survey also do not destroy the survey's probativeness. Skating Clubs alleges that the expert who conducted the survey, Doctor Edward W. Cundiff, did not (1) pretest the questionnaire to determine if the questions were confusing; (2) validate the survey by calling several interviewees and verifying their responses; and (3) initially compute the results correctly. However, Dr. Cundiff stated that he pretested the questionnaire by asking two teaching colleagues to fill them out; he felt further pretesting was unnecessary given the simplicity of the questionnaire. Dr. Cundiff also stated that he validated the survey by: (1) secretly observing the interviewers while they conducted some of the interviews; (2) asking the interviewers after the interviews whether they followed his instructions; and (3) randomly phoning several interviewees to confirm that they were interviewed; he did not attempt to confirm their answers because people often rethink their answers after the interview and offer different ones. Lastly, Dr. Cundiff stated that he recomputed the survey results accurately prior to trial. These uncontradicted statements by Dr. Cundiff show that certain safeguards existed to ensure the proper execution of the survey. Dr. Cundiff might have used better safeguards, as Skating Clubs alleges, but the ones he used are nevertheless sufficient to allow one to accord some weight to the survey results.
Finally, we emphasize that we do not decide today precisely how much weight should be given the survey results. We only uphold the district court's decision that the survey's technical deficiencies do not strip it of all probative value.