1-800 Contacts v. Lens.com Inc. et al

Filing 268

MEMORANDUM DECISION granting in part and denying in part 197 Motion to Strike the Declaration, Expert Report and Survey of Carl Degen. Signed by Judge Clark Waddoups on 12/15/2010. (jtj)

-DN 1-800 Contacts v. Lens.com Inc. et al

Doc. 268

IN THE UNITED STATES DISTRICT COURT DISTRICT OF UTAH, CENTRAL DIVISION

1-800 CONTACTS, INC., Plaintiff, v.

MEMORANDUM DECISION AND ORDER ON CARL DEGEN EVIDENCE

Case No. 2:07-cv-591 CW LENS.COM, INC. d/b/a/ LENS.COM, JUSTLENS.COM and JUSTLENSES.COM, Defendant. Judge Clark Waddoups

INTRODUCTION 1-800 Contacts, Inc. ("Plaintiff") has brought a trademark infringement action against Lens.com, Inc. ("Defendant"). A relevant factor in determining trademark infringement is whether there has been actual confusion in the marketplace. To support that actual confusion has occurred in this case, Plaintiff submitted declarations, an expert report and a survey conducted by Carl Degen ("Degen"). Defendant has moved to strike this evidence on the ground that it fails to meet the reliability requirements of Federal Rule of Evidence 702. The court grants the motion to strike the survey regarding consumer confusion and all parts of Degen's declarations and expert report that relate to the consumer-confusion survey. The remaining portions of Degen's declarations and expert report that pertain to Synovate are not stricken.

Dockets.Justia.com

SURVEY DESIGN AND RESULTS1 Degen's Background Degen has been an economist for over thirty years.2 He has a bachelor's degree in mathematics and economics and a master's degree in Economics.3 In addition, he completed the course work for a doctorate in economics.4 Both his undergraduate and graduate work included courses in statistics.5 Since 1980, he has been employed by Laurits R. Christensen Associates, Inc., "an economic research and consulting firm,"6 and is currently President of that company.7 Degen has "testified in numerous litigations involving intellectual property and other business disputes, and in postal rate cases."8 Approximately seventy-five percent of Degen's career has involved postal-related cases.9

The parties filed some of the documents under seal for this motion. The court has considered the information disclosed in this decision and concludes that it does not warrant sealing the decision.
2

1

Declaration of Carl G. Degen in Opp'n to Mot. to Strike, ¶ 3 (Dkt. No. 242, Ex. 6). Id. Id. Id. Id. ¶ 2; Carl G. Degen Resume, at 1 (Dkt. No. 242, Ex. 3, App. A). Carl G. Degen Resume, at 1 (Dkt. No. 242, Ex. 3, App. A). Id. Deposition of Carl G. Degen, 20:12­21:2 (Dkt. No. 210, Ex. C) (hereinafter "Degen -2-

3

4

5

6

7

8

9

Depo.").

As a testifying expert, most of his experience pertains to providing opinions on damages.10 Degen has experience in designing and conducting surveys. A larger part of his work, however, involves analyzing "existing survey data."11 Although Degen has been involved in survey work, until this litigation he has never designed a survey to test specifically for likelihood of confusion in a trademark case.12 Consequently, Defendant contends that Degen lacks the requisite experience to be a testifying expert. Hal Poret Background Defendant retained Hal Poret ("Poret") to review Degen's survey and conclusions. Poret is the "Vice President of Guideline, a survey research firm."13 He has "personally designed, supervised, and implemented approximately 250 consumer perception surveys," of which over 100 were trademark surveys.14 Among the trademark surveys, over 60 addressed likelihood of

confusion.15 Additionally, "[a] number of [his] surveys have specifically concerned consumer perception of trademark use in the context of internet sites, internet searches, and internet advertising."16 Poret has both a bachelor's and a master's degree in mathematics.17 He graduated

10

Id. at 33:9­17. Id. at 18:12­21. Id. at 10:10­17. Declaration of Hal Poret, ¶ 1 (Dkt. No. 210, Ex. A) (hereinafter "Poret Decl."). Id. Id. Id. Id. ¶ 2. -3-

11

12

13

14

15

16

17

from Harvard Law School and practiced in trademark litigation for 6 years.18 Poret has spoken at conferences about "how to design and conduct surveys relating to trademarks and advertising,"19 and his surveys have been accepted into evidence in both U.S. District Court litigations and "proceedings before the Trademark Trial and Appeal Board."20 Survey Design In Poret's opinion, Degen's survey is unreliable due to fatal design flaws. The survey at issue is a web-based survey comprised of four cell groups.21 The survey hypothesis was that if consumers enter the search term "1800contacts" on Google, and Lens.com appears as a sponsored link, consumers will likely be confused about whether Lens.com originates from or is affiliated with 1800Contacts.22 All survey respondents were asked a series of qualifying questions, including questions about their age and nature of their employment.23 They also were asked, "which of the following products have you purchased in the past 12 months or are you considering purchasing in the next 12 months

18

Id. Id. ¶ 3. Id. ¶ 1.

19

20

Expert Report of Carl G. Degen, ¶ 13 (Dkt. No. 210, Ex. B) (hereinafter "Degen Expert Report"). One could say the survey had five cell groups because Instrument 3 was administered twice due to an error in one of the questions. Individuals in the first group that responded to Instrument 3 were asked how frequently they had been on the Internet in the last three months. Individuals in the second group were asked how frequently they had been on the Internet in the last month. See id. ¶ 23 n.11; see also Degen Depo., 89:23­90:7 (Dkt. No. 210, Ex. C). The results from both groups were reported together for Cell Group 3.
22

21

See Deposition of Carl G. Degen, 42:18­25 (Dkt. No. 228, Ex. 67). Degen Expert Report, ¶ 14 (Dkt. No. 210, Ex. B). -4-

23

for yourself or for a member of your household. Check as many as apply."24 Among the list of possible products were contact lenses. "Only those respondents indicating they had purchased contacts within the last 12 months or were considering buying contacts within the next 12 months continued in the survey."25 "The survey was not limited to those who purchase or would consider purchasing contact lenses on the internet."26 Respondents were then asked how frequently they had been on the Internet in the last month.27 Some respondents in Cell 3 also were asked how frequently they had been on the Internet in the last three months.28 Additionally, respondents were asked whether they had used a search engine to search the Internet in the last month.29 The survey was not limited to those who used Google as their search engine in the last month. Cell Group 1 After answering these questions, the different cell groups were shown a screen shot from Google that varied by group. The first cell group viewed Instrument 1. Instrument 1instructed the respondent:

24

Id. ¶ 14 n.5. Id. ¶ 14.

25

Declaration of Hal Poret, ¶ 6 n.1 (Dkt. No. 210, Ex. A) (hereinafter "Poret Decl.") (emphasis in original).
27

26

Survey, Usage Questions (Dkt. No. 210, Ex. B, App. D). Id. Id. -5-

28

29

Imagine that you are interested in purchasing contact lenses through the Internet. Imagine that you direct your web browser to Google.com and enter the following search term: 1800contacts Click the Next button below to see the results of that search.30 The second screen showed a partial search-results page and asked the respondent to "review this search results page as you would normally do when you conduct a search."31 The page showed "organic" results and "sponsored links" results. Although "1800Contacts" typically appears as the first sponsored link whenever a person searches for "1800contacts" on Google, Degen modified the search page to delete that appearance32 because he wanted to account for defensive bidding.33 Plaintiff currently bids on its own name so that it not only appears as the first organic result, but also as the first sponsored link.34 It contends that if its competitors did not bid on its trademark, it would not have to engage in defensive bidding. Consequently, Degen designed Instrument 1 to remove defensive bidding.35 As a result, 1800Contacts only appeared under the organic results on

30

Instrument 1, First Screen (Dkt. No. 210, Ex. B, App. D). Id. at Second Screen.

31

Degen worked with Keegan & Company ("Keegen") in designing the survey screen shot. Declaration of Carl G. Degen in Opp'n to Mot. to Strike, ¶¶ 9­11 (Mar. 11, 2009) (Dkt. No. 242, Ex. 6). Keegen then entered that information into a web-based platform run by Zoomerang. Id. ¶ 11. Thus, technically Keegen edited out the information on the screen shot, but it was done under Degen's direction.
33

32

Id. ¶¶ 14­15. Id. ¶ 14.

34

As discussed below, Degen compared Cell Group 1 to Cell Group 3 because Cell Group 3 did list "1-800 Contacts" as a sponsored link. There were no appreciable differences in the level of confusion between the two cells (7.5 percent compared to 7.3 percent). Id. ¶ 16; see also Degen -6-

35

Instrument 1 and not as a sponsored link. In place of 1-800 Contacts, Degen inserted an advertisement for Lens.com as the first sponsored link. The advertisement stated: Lens.com Official Site
w w w . L e n s . c o m W e offer a wide selection of lenses at 70% off. Shop Lens.com Now.36

Below the Lens.com advertisement, four different listings appeared in the organic results. Each of them referred to 1-800 Contacts.37 After viewing the second screen, the respondent then clicked the "next" button to the third screen. The third screen had the same images as the second screen, except a large, red asterisk appeared next to the "Lens.com" advertisement.38 Respondents were asked the question: "Do you think that the link marked with the asterisk (*) originates from 1-800-CONTACTS?"39 The question appeared both above and below the screen shot. Respondents could answer "Yes," "No," or "Don't know."40 If a respondent answered "No," the person advanced to the fourth screen shot, which asked, "Do you think that the link marked with the asterisk (*) has sponsorship or approval from 1800-CONTACTS?"41 Respondents answering "Yes" on either screen three or four were not

Expert Report, ¶¶ 4­5 (Dkt. No. 210, Ex. B).
36

Instrument 1, Screen Shots 2­4 (Dkt. No. 210, Ex. B., App. D).

It is typical for multiple links to appear regarding 1-800 Contacts when the search term "1800contacts" is entered.
38

37

Instrument 1, Screen Shot 3 (Dkt. No. 210, Ex. B, App. D). Id. Id. Id. at Screen Shot 4. -7-

39

40

41

differentiated in Degen's report. Instrument 1 yielded a confusion level of 19.4 percent.42 After viewing the screen shots, respondents were then asked additional questions, including "Have you ever purchased contact lenses from 1-800-CONTACTS through the Internet?"43 Forty individuals (19.4 percent of the respondents) in Cell Group 1 answered "Yes" to this question as well.44 In the general population, only between 10 and 11 percent purchase contacts through the Internet, and not all of those individuals purchase from 1-800-Contacts.45 Cell Group 2 and 3 Cell Group 2 viewed Instrument 2. It was a control group and thus, the first screen informed the respondents that the search term was "contact lenses" rather than "1800contacts."46 By substituting the search term, Degen tested for how many respondents would still say that Lens.com originated from or was affiliated with 1-800-Contacts even when the search term did not reference "1800contacts." This was done to determine the level of "noise." Noise is the "tendency of survey respondents to give a particular answer for reasons other than the hypothesis that you're studying."47 Except for changing the search term and dropping off the last line at the bottom of the screen, Instrument 2 was identical to Instrument 1. Instrument 2 yielded a confusion level of 11.9 percent.48

42

Degen Expert Report, ¶ 21 (Dkt. No. 210, Ex. B). Survey, Background Question 1 (Dkt. No. 210, Ex. B, App. D). See id. Historical Telenation Survey Results (Synovate) (Dkt. No. 181, Ex. 1, Attach. A). Instrument 2, First Screen (Dkt. No. 210, Ex. B, App. D). Deposition of Hal Poret, 101:8­13 (Dkt. No. 242, Ex. 2) (hereinafter "Poret Depo."). Degen Expert Report, ¶ 22 (Dkt. No. 210, Ex. B). -8-

43

44

45

46

47

48

Cell Group 3 viewed Instrument 3. The search term for Instrument 3 reverted back to "1800contacts."49 Degen changed the sponsored links, however, to include both 1-800-Contacts and Lens.com at the top of the screen. The first sponsored link said, 1800CONTACTS.com www.1800contacts.com Need contact lenses? Buy online in just 5 minutes. We make it simple.®50 The second link was for Lens.com and it contained the same advertisement as that stated above for Instrument 1. Despite the addition of a second sponsored link, the size of the screen shot remained the same because Degen deleted the fourth organic search result in full and also the bottom line of the third organic result. Otherwise, Instrument 1 and Instrument 3 were the same. Instrument 3 yielded a confusion level of 19.2 percent.51 Cell Group 4 Cell Group 4 viewed Instrument 4, which differed substantially from the other instruments. Respondents again were informed the search term was "1800contacts."52 On the second screen shot, however, no sponsored link appeared at the top of the page. Four organic search results appeared in the left column. They were the same as those in Instrument 1, except the fourth result contained additional text about 1-800 Contacts. The right column contained three sponsored links, with the following language:

49

Instrument 3, First Screen (Dkt. No. 210, Ex. B, App. D). Instrument 3, Screen Shots 2­4 (Dkt. No. 210, Ex. B, App. D). Degen Expert Report, ¶ 23 (Dkt. No. 210, Ex. B). Instrument 4, First Screen (Dkt. No. 210, Ex. B, App. D). -9-

50

51

52

ContactLens.com 75% Off Up to 75% off Retail Price! Free Shipping on Orders Over $89 www.ContactLens.com 1-800 Contacts Simple online ordering of lenses. Compare our prices and save! www.JustLenses.com Discount Contact Lenses Buy online and save up to 50% on your contacts www.discountedcontactlense.com53 On the third screen shot, the same images appeared except, a large letter "L" was placed next to the first sponsored link, a large letter "M" was placed next to the middle sponsored link, and a large letter "N" was placed next to the last sponsored link.54 Respondents were directed, "Please indicate the corresponding letter of all lettered links, if any, that you think originate from 1-800CONTACTS or have sponsorship or approval from 1-800-CONTACTS. Select all that apply."55 This directive appeared both above and below the screen shot. Respondents could answer "L," "M," "N," "All the links identified," or "None of the links identified."56 If a respondent selected "M," Degen concluded the advertisement caused consumer confusion.57 Instrument 4 yielded a confusion

53

Id. at Screen Shot 2 (emphasis in the original). Id. at Screen Shot 3. Id. Id. Degen Expert Report, ¶ 31 (Dkt. No. 210, Ex. B). -10-

54

55

56

57

level of 38.4 percent for link "M".58 It also yielded a confusion level of 24.4 percent for link "L" and 22.7 percent for link "N."59 There was no specific control group for this instrument. The following chart summarizes the data for the four cell groups.
Cell 1 C e ll 2 C o n t r o l Group 219 c o n ta c t lenses L e n s .c o m - first s p o n s o r e d listing C e ll 3 C e ll 4

S a m p le Size S e a r c h Term S p o n s o r e d Link

206 1 8 0 0 c o n ta c ts L e n s . c o m - first s p o n s o r e d listing

4 1 2 (two tests) 1 8 0 0 c o n ta c ts 1 -8 0 0 -C o n ta c ts - first s p o n s o r e d listing, followed b y Lens.com as the second s p o n s o r e d listing L e n s .c o m Official Site

218 1 8 0 0 c o n ta c ts T h r e e sponsored lin k s in right h a n d column

A d v e r tis e m e n t H e a d e r of R e le v a n t Link C o n fu s io n ( " Y e s " response)

L e n s . c o m Official S ite

L e n s .c o m Official S ite

1 -8 0 0 Contacts

1 9 .4 %

1 1 .9 %

1 9 .2 %

2 4 .5 % - L 3 8 .4 % - M 2 2 .7 % - N unkno wn 17%

N e t Confusion E v e r Purchased C o n ta c ts from 18 0 0 Contacts on I n te rn e t?

7 .5 % 19%

n /a - control group 20%

7 .3 % 22%

"Net Confusion" was calculated by subtracting the control group's confusion score from a cell group's confusion score. This eliminated the "noise" that existed in Cell Groups 1 and 3. Once noise was eliminated, the net confusion rate for Cell Groups 1 and 3 dropped to 7.5 percent and 7.3 percent respectively. The "noise" that existed in Cell Group 4 cannot be determined based on the data Degen presented in his expert report.

58

Id. Id. -11-

59

ANALYSIS I. STANDARD FOR ADMISSIBILITY OF EXPERT TESTIMONY Rule 702 of the Federal Rules of Evidence specifies that if an expert's opinion "will assist the trier of fact to understand the evidence or to determine a fact in issue," that expert may testify "if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case."60 The Rules of Evidence have a "liberal thrust,"61 such that there is a "strong and undeniable preference for admitting any evidence having some potential for assisting the trier of fact."62 Nevertheless, the court still retains a gatekeeper role to ensure that "expert testimony is both reliable and relevant" before it is admitted.63 When determining whether an expert opinion is reliable, the court "must assess the reasoning and methodology underlying the expert's opinion."64 Although Defendant moved to strike the Degen evidence, it is Plaintiff that bears the burden of proving that Degen's survey, report, and declarations are admissible because it is the proponent of the evidence.65

60

Fed. R. Evid. 702. Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 588 (1993). United States v. Velasquez, 64 F.3d 844, 849 (3d Cir. 1995) (quotation marks and citation

61

62

omitted). United States v. Rodriguez-Felix, 450 F.3d 1117, 1122 (10th Cir. 2006) (citing Fed. R. Evid. 702).
64 63

Id. at 1123 (quotation marks and citations omitted).

Ralston v. Smith & Nephew Richards, Inc., 275 F.3d 965, 970 n.4 (10th Cir. 2001) (citations omitted). -12-

65

II.

EXPERTS' QUALIFICATIONS A. Degen's Qualifications

Defendant moves to strike Degen's survey on the basis that he does not qualify as an expert. Although Degen has had course work in statistics and has designed and administered surveys, he has never designed a survey to address likelihood of confusion. Rather, many of his surveys pertained to gathering objective data or information about damages. Consequently, Plaintiff has failed to demonstrate that Degen is an expert in consumer-confusion surveys, which impacts his credibility. As stated above, however, the rules of evidence have a liberal thrust. Thus, "rejection of expert testimony has been the exception rather than the rule."66 Although Degen is not an expert in consumer-confusion surveys, Degen does have some experience designing and administering surveys. Such experience is relevant to some degree because certain rules in survey design and administration exist across disciplines. The court therefore concludes that while Degen's relevant experience is not extensive, it is sufficient for him to qualify as an expert in this matter. B. Poret's Qualifications

In its memorandum opposing Defendant's motion to strike, Plaintiff contends that portions of Poret's declaration should be stricken. It contends that Poret is not qualified to opine on the survey's proper universe because he lacks the proper education and understanding of the relevant market. Hence, it argues that Poret's opinion on the proper universe should be excluded. Offering these arguments in an opposition memorandum falls far short of Plaintiff's

Ruff v. Ensign-Bickford Indus., Inc., 171 F. Supp. 2d 1226, 1232 (D. Utah 2001) (citing Goebel v. Denver & Rio Grand W. R.R. Co., 215 F.3d 1083, 1089 (10th Cir. 2000)) (other citation omitted). -13-

66

obligation to file both a motion and a supporting memorandum if it intends to move the court to strike a declaration.67 "All attorneys practicing before this court, whether admitted as members of the bar of this court, admitted pro hac vice, or otherwise . . . are governed by and must comply with the rules of practice adopted by this court . . . ."68 Moreover, Plaintiff's objections to Poret's qualifications are not well-taken. He has extensive experience in consumer survey design and is far more acquainted about the proper universe for such surveys than Degen. Although substantial time was spent in Poret's deposition focusing on the documents and areas he did not review, the court concurs that Plaintiff's marketing and advertising strategy has little to do with whether consumers are confused by a Lens.com advertisement on Google. Even taking as true that Plaintiff expends significant effort trying to attract new customers to purchase contacts via the internet and that such customers are more apt to use a search engine to find 1-800 Contacts, it does not follow that Plaintiff's marketing and advertising efforts form the basis for this survey's proper universe. Accordingly, the court declines to exclude Poret's declaration regarding the proper universe.69 III. SURVEY DESIGN FLAWS Defendant argues that Degen's survey is so riddled with flaws that it cannot pass the

See DUCivR 7-1(a)­(b). Compliance with the rule is important not simply for the sake of enforcing the rule, but because a properly filed motion allows the court to assess its merits with a proper focus on the support provided and with a fair opportunity for the opposition to address the issues raised. Plaintiff's argument to strike fails to meet these standards.
68

67

DUCivR 83-1.1(g).

Plaintiff also objected to other portions of Poret's declaration, which will be addressed later in this decision. -14-

69

"reliability" prong for admissibility. To satisfy reliability standards, "a survey offered to establish the likelihood of consumer confusion must have been fairly prepared and its results directed to the relevant issues."70 "Two hallmarks of a good survey include the selection of an appropriate universe of respondents and the use of non-leading questions."71 Defendant asserts the survey has neither a proper universe nor non-leading questions. A. Survey Universe

Defendant contends that the survey's universe was over-inclusive. "A `universe' is that segment of the population whose perceptions and state of mind are relevant to the issues in the case."72 "Identification of the proper universe is recognized uniformly as a key element in the development of a survey."73 "The definition of the relevant population is crucial because there may be systematic differences in the responses of members of the population and nonmembers."74 At times, an over-inclusive universe can be corrected if "a sufficiently large (and representative) subset of respondents in the survey was drawn from the appropriate universe."75 "If the relevant subset

LG Elecs. U.S.A., Inc. v. Whirlpool Corp., 661 F. Supp. 2d 940, 951 (N.D. Ill. 2009) (quotation marks and citations omitted). Bobak Sausage Co. v. A & J Seven Bridges, Inc., No. 07-C-4718, 2010 U.S. Dist. LEXIS 40737, at *22 (N.D. Ill. 2010). Citizens Fin. Group, Inc. v. Citizens Nat'l Bank, 383 F.3d 110, 118­19 (3rd Cir. 2004) (quotation marks and citation omitted). Shari Seidman Diamond, Reference Guide on Survey Research, REFERENCE MANUAL ON SCIE N T IFIC EVIDENCE, 239 n. 41 (Fed. Judicial Ctr. 2d ed. 2000) (Dkt. No. 242, Ex. 1) (hereinafter "Survey Research Manual").
74 73 72 71

70

Id. at 240. Id. at 242. -15-

75

cannot be identified, however, an overbroad universe will reduce the value of the survey."76 Plaintiff argues that the universe was proper because it only included individuals who had purchased contacts within the prior twelve months or who intended to purchase contacts within the next twelve months. Because Plaintiff's marketing and advertising targets all contact lens users and the survey was limited to contact lens purchasers, it contends it got the universe right. The court disagrees. "[A] survey to test likelihood of confusion must attempt to replicate the thought processes of consumers encountering the disputed mark or marks as they would in the marketplace."77 The survey was not about who Plaintiff targets as customers. It was about whether consumers who purchase contacts on the Internet are likely to be confused by a Lens.com sponsored link. Only those who have purchased or are likely to purchase contacts on the Internet would see Lens.com's sponsored link in the actual market. Accordingly, that was the relevant universe. Yet, the survey questions did not limit respondents only to those who purchased contacts on the Internet.78 The court therefore concludes that Degen used an over-inclusive universe. Although about 20 percent of the total survey respondents did answer that they had purchased contacts on the Internet from 1-800 Contacts, the court questions whether the size of the subset is sufficient to allow meaningful data to be derived from it. Even if the subset's size is sufficient, no

76

Id.

Bobak Sausage Co., 2010 U.S. Dist. LEXIS 40737, at *14 (quotation marks and citations omitted) (emphasis added). The survey also asked respondents to review the search-results page as they "normally do" when conducting a search on the Internet. This, too, shows the appropriate marketplace for the survey was the Internet and not all possible outlets where contacts may be purchased. -1678

77

data was presented in Degen's report to show how these particular respondents answered the questions presented on the screen shots. Accordingly, no analysis can be made for that subset. The court therefore concludes that the survey universe was improper and cannot be corrected by looking only at a subset. B. Improper Questions

Defendant further contends that the survey questions were improper because they were closeended and leading. When "evaluating the objectivity of a survey," a court must consider "the degree of suggestiveness of each question."79 Close-ended questions are not inherently biased or leading.80 Indeed, they can be "suitable for assessing choices between well-identified options" and "may remind respondents of options that they would not otherwise consider or which simply do not come to mind as easily."81 "The advantage of open-ended questions," however, "is that they give the respondent fewer hints about the answer that is expected or preferred."82 1. Instruments 1, 2, and 3

In this case, survey respondents were presented close-ended questions. On the first screen shot for Instruments 1 and 3, respondents were told the search term was "1800contacts." They also were told that the next screen would show "the results of that search." On the third screen shot for Instruments 1, 2, and 3, respondents were presented twice with the question: "Do you think that the

79

LG Elecs. U.S.A., Inc., 661 F. Supp. 2d at 954 (quotation marks and citations omitted). Id. Id. at 954­55 (quotation marks and citations omitted). Survey Research Manual, supra note 73, at 252. -17-

80

81

82

link marked with the asterisk (*) originates from 1-800-CONTACTS?" Survey respondents had the option of choosing "yes," "no," or "don't know." Neither the question nor the options presented the respondents with any name other than 1-800 Contacts. Thus, there was no opportunity to remind respondents of other options they might not have considered. An open-ended question was not required to create a valid survey question. Indeed, an openended question may have made interpreting the survey results more difficult. The problem, however, is not the failure to use an open-ended question, but the failure to design a close-ended question that offers exhaustive alternative responses. Here, the form of the question strongly suggested the response.83 A close-ended question, to maintain the objectivity of the survey, must allow the respondent the unbiased choice from a reasonable list of choices.84 A close-ended question which allowed the respondent multiple choices or a scale from which to chose an answer may have avoided the suggestiveness inherent in the questions that destroy the reliability of this survey.85 Moreover, there is a problem with the clarity of the question. "When unclear questions are included in a survey, they may threaten the validity of the survey by systematically distorting responses if respondents are mislead in a particular direction, or by inflating random error if

"1-800 Contacts" also appeared in bold a number of times on the screen under the sponsored and organic search results. This further highlighted the phrase to survey respondents. At times, the choice of "yes," "no," or "don't know" may be exhaustive. In this case, however, it did not present a reasonable list of choices because the only name suggested to the survey respondents was "1800Contacts." For example, had the question and list been designed also to include such choices as "Lens.com," "ContactLens.com" or "DiscountedContactLense," it would have lessened the suggestiveness of the answer. See e.g., Survey Research Manual, supra note 73, at 252­53; Bobak Sausage Co., 2010 U.S. Dist. LEXIS 40737, at *17 (giving respondents an option to choose from a range of answers was a useful feature). -1885 84

83

respondents guess because they do not understand the question."86 Notably, "[i]f the crucial question is sufficiently ambiguous or unclear, it may be the basis for rejecting the survey."87 In this case, respondents were told the search term was "1800contacts" and that the next screen would show the results of that search. They were then asked if the highlighted link "originates from 1-800CONTACTS." Given the context in which this question was asked, the meaning of "originates from" was ambiguous. Certainly, the link originated from the search term "1800contacts." All search results on the page originated from the search term. Thus, a respondent could have answered "yes" to the question without understanding that the question did not refer to the search term but to the company paying for the advertisement.88 Degen performed no pre-tests to find out "unexpected question interpretations" before the survey was conducted.89 Nothing in the survey allowed the surveyor to test for or correct interpretation errors. The problem with the survey question is further exacerbated by the lack of data reported in Degen's expert report. Although it provided a specific breakdown of responses for the background, usage, and qualifying questions, no such information was provided for the responses to the screen

86

Survey Research Manual, supra note 73, at 248. Id.

87

See e.g., PBM Products, LLC v. Mead Johnson & Co., No. 3:09-cv-269, 2010 U.S. Dist. LEXIS 18404, at *11­13 (E.D. Va. Mar. 2, 2010) (concluding that the failure to define a material term in a survey was fatal). As noted above, the survey was conducted through Zoomerang's platform. Zoomerang itself encourages surveyors to "pre-test your survey with a few members of your target audience and/or co-workers to find glitches and unexpected question interpretations." MarketTools, Inc., Zoomerang, 10 Tips to Improve Your Surveys, 3 (2005) (Dkt. No. 242, Ex. 6, Attach. 2) (emphasis added). -1989

88

shots. Instead, only combined totals were presented on each of the instruments. Accordingly, the court has no data about how many "yes" responses occurred as a result of screen shot three (originates from) versus screen shot four (sponsorship) for Instruments 1, 2, and 3. In his deposition, Degen asserted he chose the word "originates" because that is "the word that appears in the Lanham Act with respect to confusion."90 He further believed it was

straightforward.91 Merely because a word appears in statute does not make it the ideal word for a survey question. While statutory language may be "well-defined for attorneys" or experts, it may not be "meaningful for laypersons."92 Because "originates from" can have multiple meanings to a layperson, particularly in the context in which it was used in the survey, the court concludes the question was ambiguous. 2. Instrument 4

The same problems exist for Instrument 4. Survey respondents were asked: "Please indicate the corresponding letter of all lettered links, if any, that you think originate from 1-800-CONTACTS or have sponsorship or approval from 1-800-CONTACTS. Select all that apply."93 Not only does the question contain the same ambiguous phrase, but it combines two questions in one. It is therefore impossible to know which question elicited the particular responses that were given. Without knowing which question elicited a response, no meaningful analysis can be made because

90

Degen Depo., 97:22­24 (Dkt. No. 210, Ex. C). Id. at 98:7­20. Survey Research Manual, supra note 73, at 248. Instrument 4, Third Screen (Dkt. No. 210, Ex. B, App. D). -20-

91

92

93

results derived from the ambiguous phrase cannot be factored out. C. Other Design Flaws and Undue Prejudice

Defendant also asserts the survey failed to approximate actual market conditions and that the cell groups were uncontrolled, which further skewed the data. While Defendant may be correct that these additional design flaws exist as well, the court does not address them here because it is apparent that Degen's survey cannot survive the motion to strike. Although technical flaws typically go to weight rather than admissibility, the flaws in this survey's universe and crucial questions are so substantial as to render the survey unreliable. The problem is exacerbated by Degen's failure to design the survey so as to allow subclasses to be separately examined to test the validity of the information. For example, the threshold question was compound without any ability to determine which respondents had actually used or would use the Internet to purchase contact lenses. The court therefore concludes that Plaintiff has failed in its burden of establishing that Degen's survey is admissible under Rule 702. III. DEGEN'S DECLARATIONS AND EXPERT REPORT Besides moving to strike Degen's survey, Defendant also has moved to strike Degen's declaration and Degen's expert report. Degen has submitted several declarations pertaining to the consumer-confusion survey. Because the court has stricken the survey, the court necessarily strikes all portions of Degen's declarations and Degen's expert report that pertain to the consumerconfusion survey. Plaintiff correctly asserts, however, that Degen also has opined about a survey conducted by Synovate. Defendant does not address the Synovate survey in its memorandum. Accordingly, the

-21-

court declines to strike those portions of Degen's declarations or expert report that address the Synovate survey. IV. REMAINING PORTIONS OF PORET'S DECLARATION In its memorandum opposing Defendant's motion to strike, Plaintiff also contends that other portions of Poret's declaration should be stricken. Specifically, Plaintiff objects to paragraphs 18­19, 28­31, and 67­72 because Poret addresses the Synovate survey. Plaintiff objects to paragraphs 63­65 because they allegedly improperly address issues pertaining to Cell Group 4. Finally, Plaintiff objects to paragraphs 21, 33­36, and 73 because they allegedly state legal conclusions. Again, these arguments are not well-taken because Plaintiff failed to move on them properly. Moreover, the issue about the Synovate survey has been resolved by Paragraph III above, and is therefore moot. With respect to Cell Group 4, the court did not reach the issue of whether the survey utilized proper control-group methodology. Consequently, the court did not rely on that portion of Poret's declaration and Plaintiff' objection is moot. Regarding Plaintiff remaining objections, paragraph 73 does not render a legal conclusion, and the court did not rely on paragraphs 21 and 33­36 in forming its opinion. Accordingly, Plaintiff' objections are either erroneous or moot. CONCLUSION For the reasons stated above, the GRANTS IN PART AND DENIES IN PART Lens.com's Motion to Strike the Declaration, Expert Report, and Survey of Carl Degen.94 The court strikes Degen's consumer-confusion survey and all parts of his expert report and declarations that relate to

94

Docket No. 197. -22-

that survey. The court declines to strike those portions of Degen's expert report and declarations that relate to the Synovate survey. DATED this 15th day of December, 2010. BY THE COURT:

____________________________________ Clark Waddoups United States District Judge

-23-