Case Information
*1 Before P OSNER , R OVNER , H AMILTON , Circuit Judges . P OSNER Circuit Judge
. This infringement suit (see Lanham Act, U.S.C. §§ et seq .) brought against Old Country Store (we dis regard parties). district judge granted injunction against sale Old *2 ‐ Country Store of food products grocery stores under name Cracker Barrel, a registered trademark of Kraft. To prevent confusion (an especially apt goal in a trademark case), we’ll call Old Country Store “CBOCS,” as do parties.
Kraft a well ‐ known manufacturer of food products sold in grocery stores. Its include a wide variety packaged cheeses, number them under trade marked “Cracker Barrel” label. Kraft been selling cheese in grocery stores under that name for more than half cen tury. Thousands carry Kraft bear ing that label. Kraft does sell non ‐ cheese under Barrel. well known chain low price restaurants (it
opened its first restaurant 1969), number last count, many just off major highways. Upon learning recently planned variety food prod ucts (not including cheese, however), packaged hams, logo, “Cracker Old Country Store ” (the last three words are smaller type logo), Kraft filed suit. It claims many will confused similarity logos think labeled products, result they dissatisfied product they blame Kraft. acknowledges does entitle
owner prevent all uses even identical marks. “It would hard, example, seller steam shovel find ground complaint use his trade mark lipstick.” L.E. Waterman Co. Gordon (2d 1934) (L. Hand, J.). And likewise iden *3 3 tical marks used on sold through different types sales outlet might cause no confusion—indeed does question CBOCS’s right to food issue CBOCS’s restaurants, CBOCS’s small “country stores” adjoin restau rants, mail order or on Web. It objects only sale stores. The district judge found likelihood confusion, resulting harm Kraft, from CBOCS’s selling through outlets sufficient warrant grant preliminary injunction. These factual determinations, bind us unless clearly erroneous, Eli Lilly & Co. v. Natural Answers, Inc. , F.3d (7th Cir. 2000); Scan dia Down Corp. Euroquilt, 1427–28 (7th 1985)—a reinforcing consideration being need expeditious determination whether order preliminary relief. The district judge must act certain haste, must hesitate nitpick his findings casually remand further proceedings bound cause additional delay. moved preliminary injunction on March year; was granted July 1; it now November. grant preliminary injunction followed extensive discovery, presentation expert evidence, some live testimo ny. Of allegedly infringing products, only spiral hams had been shipped before injunction was issued; then had all.
Below, copied CBOCS’s website, picture logo appeared shipped gro cery stores.
Up close least, looks different label “Cracker Barrel” appears on Kraft’s cheeses. Yet even cheese ham (or products) were displayed side side store, would make shopper notice difference between labels, words “Cracker Barrel” on both labels—and much larger type than “Old Country Store” CBOCS’s la bel—might lead shopper think both prod ucts.
Most cheese must know it’s product, “Kraft” typically though invariably appears label, follow ing picture:
Kraft concerned the potential confusion shop ‐ pers at the 16,000 so (or similar retail enti ‐ ties) Cracker Barrel cheeses, the also carry food products logo (not only ham but delicatessen meats, bacon, sausages, jerky, meat glazes, baking mixes, coating mixes, oatmeal, grits, gra vies—all are by CBOCS). Were cheeses meats exhibited side side shelf, difference appearance logos two brands might said lead some consumers think they were made different companies—but might lead others think opposite, since different products same manufacturer are often exhibited together. If hand food are dif ferent locations store, some consumers might forget difference between logos think all products. Cf. Ty, Jones Group, Inc. 898–99 (7th 2001). Even savvy might fooled, because they know producers often vary ap pearance trademarks.
It’s fact parties’ trade names simi lar decisive, nor even fact (low cost packaged items). It those similari *6 ties coupled fact that, if prevails suit, products confusingly similar trade names be sold through same distribution channel—grocery stores, often same grocery stores—and advertised together. (In brief period before injunc tion was issued, hams were sold grocery stores, an online ad for Cracker Barrel Sliced Spiral Ham by coupons firm provided link to coupon Kraft’s Cracker Barrel cheeses.) The competing would be likely appear same store circulars. Such similari ties overlap would increase likelihood confusion detrimental Kraft. See CAE, Inc. v. Clean Air En gineering, Inc ., F.3d (7th 2001); Ty, Jones Group, Inc ., supra 900–01.
Still another reason expect confusion both most meat licensed sale grocery inexpensive— $5. Generally only very cost conscious consumers apt scrutinize carefully labels less expensive items store. Familiarity have made name salient shoppers, product bearing might attributed even close scrutiny label would suggest product might well different origin.
If significant number confused names thought CBOCS’s were made Kraft, could badly hurt. A trademark’s value saving search costs made possible information conveys about quality trademark own er’s brand. brand’s reputation quality depends owner’s expenditures product quality quality *7 ‐ 7 control, service, advertising, so on. Once reputation is created, firm will obtain greater profits because repeat purchases word mouth endorsements will add to sales because consumers willing to pay a higher price exchange for a savings search costs an assur ance consistent quality. These benefits depend on firm’s ability to maintain consistent quality. When a brand’s quality inconsistent, learn trademark does enable to predict their future con sumption experiences from their past ones. The trademark does then reduce search costs. They become unwill ing pay more branded than for unbranded good, firm no longer earns a sufficient return expenditures promoting justify them. particular danger CBOCS’s being al lowed products through same outlets trade name confusingly Kraft’s “Cracker Barrel” trade CBOCS’s inferior respect what consumer expects—if consumer bad experience product blames Kraft, thinking it producer—Kraft’s sales cheeses decline; consumer who thinks makes bad hams may decide it probably makes bad well. See, e.g., Ty v. Perryman , F.3d (7th Cir. 2002); Omega Importing Corp. Petri Kine Cam era Co. (2d 1971) (Friendly, C.J.).
Granted, there interest tension interest avoiding confusion. Consumers benefit having variety choose among. wants offer shoppers sells restaurants. Think Starbucks—a notable example chain *8 ‐ of restaurants (coffee houses) that sells its major food prod ‐ uct (coffee) in grocery stores as well. The preliminary injunc ‐ tion prevents CBOCS from doing that pending resolution case. But CBOCS and utilizes an alternative channel to consumers, outside restaurant channel—an alterna tive channel ever greater significance in electronic age: Web. CBOCS’s restaurants invite their legions cus tomers visit CBOCS website, which displays pictures hams other food products that it sells and links buying food CBOCS online. Some foods are in CBOCS’s “country stores” adjoining res taurants. Yet doubtless thinks it reach additional consumers placing food products in grocery stores— else it wouldn’t issued licenses invited this suit (which anticipated).
So competition in food will be harmed prevails suit, extent sale confuses result impairs sales reasons having nothing do with any product quality problems Kraft. But competition will helped extent are able offer customers an additional product line. weighing balancing these compet ing interests precision feasible undertakings preliminary injunction proceeding, probably full trial either. Imponderables likely dominate.
About all feasible preliminary injunction stage judge estimate likelihood plaintiff prevail full trial parties harmed more ruling, granting denying injunction, favor party, com
*9
13 2559 9 bine these findings the manner suggested in cases as
Abbott Laboratories v. Mead Johnson & Co
.,
But grant preliminary injunction be proper, harm plaintiff must be judged irreparable—
meaning fully compensable or avoidable issuance final judgment (whether damages judgment or per manent injunction, both) plaintiff’s favor. See, e.g.,
Abbott Laboratories v. Mead Johnson & Co
.,
supra
,
Consistent analysis, if plaintiff strong likelihood prevailing full trial, costs him if injunction denied least great as costs defendant granted, plaintiff’s costs could fully recouped him final judgment his favor, injunction should issued. Roland Machin ery Co. v. Dresser Industries, Inc ., F.2d 380, 387–88 (7th Cir. 1984); Omega Satellite Products Co. v. City Indianapolis , (7th 1982); 11A Charles Alan Wright, Ar thur R. Miller & Mary K. Kane, Federal Practice Procedure § 2948.3, pp. 202–11 (3d ed. 2013); cf. Weinberger Romero Barcelo U.S. (1982). That seems situation here, given district judge’s findings. likelihood *10 confusion seems substantial and risk of loss of valuable goodwill and control therefore palpable. And as emphasized in Abbott Laboratories Processed Plastics opinions cited earlier, irreparable harm is especially likely trademark case because of difficulty of quantifying effect on brand of nontrivial period of consumer confusion (and interval between filing infringement complaint final judgment is sure trivial). And on side ledger, there is no ‐ formation how many new customers expect obtain selling through stores, given it al ‐ ready sells its its country its website. So there no basis concluding it losing heavily as result being able through until unless obtains final judgment fa ‐ vor.
So grant injunction must af ‐ firmed. But mainly future reference want say something about survey presented support claim confusion. Consumer surveys con ‐ ducted party hired expert witnesses prone bias. There such wide choice survey designs, none fool proof, involving issues as sample selection size, presentation allegedly confusing con sumers involved survey, phrasing questions way intended elicit surveyor’s desired re sponse—confusion lack thereof—from survey re spondents. See Robert H. Thornburg, “Trademark Surveys: Development Computer Based Survey Methods,” John Marshall Rev. Intellectual Property L. (2005); Michael Rappeport, “Litigation Surveys—Social ‘Science’ Evi dence,” Trademark Rep . 960–61 (2002); Jacob Jacoby, *11 2559 11 “Experimental Design and Selection Controls Trademark and Deceptive Advertising Surveys,” Trade ‐ mark Rep . 890, (2002); see Joseph Sanders, “Science, Law, and Expert Witness,” Law & Contemp. Probs. , Winter pp. 63, 73–75. Among problems identified academic literature following: when a con ‐ sumer a survey respondent, this changes normal envi ‐ ronment he or she encounters, compares, and re acts trademarks; a survey that produces results contrary interest party that sponsored survey may suppressed thus never become part trial record; expert witnesses who conduct surveys aid liti gation likely biased favor party hired paying them, usually generously. All too often “ex perts abandon objectivity become advocates side hired them.” Id . 75.
Of course, judges jurors their own biases blind spots. As Judge Jerome Frank noted many years ago, dissenting pair cases Seventeen maga zine had brought against makers “Miss Seventeen” girdles, “as neither trial judge nor member court (or resembles) teen age girl or mother or sister girl, our judicial notice apparatus work well unless feed information directly obtained ‘teen agers ʹ from female relatives accustomed shop them.” Triangle Publications, Rohrlich (2d 1948). And judge’s finding confusion was was “nothing but surmise, conjec ture, guess.” Id.
Nevertheless it’s clear caution required screening proposed experts surveys. Kraft’s *12 expert this case was Hal Poret, an experienced survey re ‐ searcher, won’t hold against him that he appears to basically a professional expert witness. See www.pli.edu/Content/Faculty/Hal_Poret/_/N 4oZ1z138h0? ID=PE830174 (visited Nov. 2013). Poret was able to ob tain a random at least representative sample Amer ican whole ham products, email them photo graphs sliced spiral ham, ask them email whether company that makes ham also makes products—and if what products. About a quarter respondents said cheese. It’s difficult know what make this. respondents may have assumed that company logo that does specify particular product doesn’t make just sliced spiral ham. So now they have guess what else company would make. Well, maybe cheese.
Poret showed control group respondents essen tially same ham, but made by Smithfield—and none these respondents said that Smithfield makes cheese. Poret inferred “Cracker Barrel” ham shown respondents had triggered their rec ollection cheese, rather than word “ham” being trigger. That plausible, but relevance obscure. Kraft’s concern people will think made by but they think ham made Kraft, event they bad experience ham they’ll blame Kraft.
Also it’s very difficult compare people’s reactions photographs shown online survey company reactions they looking *13 store trying decide whether buy. The contexts radically different, the stakes much higher when actual shopping decisions have be made (because means parting with money), which may influence responses.
In some cases an attractive alternative survey might be the use of statistical data determine the effect of the al legedly infringing logo. Suppose before this suit was filed, CBOCS had been sold time in number grocery stores. Probably in some Kraft Cracker Barrel cheese would have been displayed side by side with CBOCS hams plus meat comparable prices, while other stores hams would have been displayed different areas store, still other would have carried CBOCS hams but Kraft Cracker Barrel cheese. By examining “lift” (greater sales) if hams obtain proximity label, an expert witness might able estimate extent consumer confusion. great er lift (and hence greater confusion) greater likelihood consumer’s blaming as sup posed maker hams has bad experience hams. Such study would have been feasible case, however, given grant injunction, kept hams logo out now. Nor such confidence reliability such study we would think an adequate basis refusing grant pre liminary injunctions cases.
We imagine types expert testimony might illuminating case this—testimony experts retail about buying habits *14 psychology of of inexpensive products. “Although the ordinary consumer’s mindset is central law and policy, neither courts nor commentators have made any serious attempt develop framework understanding the conditions may affect the attention expected given particular purchase. Some of the classic judicial descriptions cast the ordinary consumer ‘ignorant … unthinking and … credulous’ or ‘hasty, heedless and easily deceived.’ In cases, courts have bristled at ‘claimed asininity’ of buying public, suggesting instead average buyer is ‘neither savant nor dolt,’ but one who ‘lacks special competency reference matter hand but exercises normal measure of layman’s common sense judg ment.’ For most part, however, debate vacuous war words, uninformed careful theoretical model ing consumer psychology empirical study consumer behavior.” Thomas R. Lee, Glenn L. Christensen & Eric D. DeRosia, “Trademarks, Consumer Psychology, So phisticated Consumer,” Emory L.J . 575–76 (2008) (footnotes deleted).
We doubts about probative significance Poret survey. But similarity logos products, channels distribution (and advertising over lap) allowed through under logo, availability company alternatives reaching large public logo, provide adequate support issuance injunction. judgment therefore
A FFIRMED .