GEICO
269
UNITED STATES DISTRICT COURT
FOR THE EASTERN DISTRICT OF VIRGINIA
ALEXANDRIA DIVISION
GOVERNMENT EMPLOYEES . Civil Action No. 1:04cv507
INSURANCE COMPANY, .
.
Plaintiff, .
.
vs. . Alexandria, Virginia
. December 15, 2004
GOOGLE, INC., . 10:00 a.m.
.
Defendant. .
.
. . . . . . . . . . .
TRANSCRIPT OF BENCH TRIAL
BEFORE THE HONORABLE LEONIE M. BRINKEMA
UNITED STATES DISTRICT JUDGE
VOLUME II
APPEARANCES:
FOR THE PLAINTIFF: CHARLES D. OSSOLA, ESQ.
CHRISTOPHER WINTERS, ESQ.
MICHAEL J. ALLAN, ESQ.
ROBERTA L. HORTON, ESQ.
Arnold & Porter LLP
555 Twelfth Street, N.W.
Washington, D.C. 20004
and
JOHN F. ANDERSON, ESQ.
Troutman Sanders LLP
1660 International Drive, Suite 600
McLean, VA 22102
and
JONATHAN L. SHAFNER, ESQ.
Government Employees Companies
One Geico Plaza
Washington, D.C. 20076
(APPEARANCES CONT'D. ON FOLLOWING PAGE)
(Pages 269 - 295)
COMPUTERIZED TRANSCRIPTION OF STENOGRAPHIC NOTES
270
1 FOR THE DEFENDANT: MICHAEL H. PAGE, ESQ.
JOSHUA H. LERNER, ESQ.
2 ANJALI S. SAKARIA, ESQ.
Keker & Van Nest, LLP
3 710 Sansome Street
San Francisco, CA 94111-1704
4 and
MELANIE D. COATES, ESQ.
5 Wilmer Cutler Pickering Hale and
Dorr LLP
6 1600 Tysons Boulevard, Suite 1000
McLean, VA 22102
7 and
MICHAEL S. KWUN, ESQ.
8 Google
1600 Amphitheatre
9 Mountain View, CA 94043
10
ALSO PRESENT: RHONDA L. ANDREW
11 DAVID GORDON
12
OFFICIAL COURT REPORTER: ANNELIESE J. THOMSON, RDR, CRR
13 U.S. District Court, Fifth Floor
401 Courthouse Square
14 Alexandria, VA 22314
(703)299-8595
15
16
17
18
19
20
21
22
23
24
25
271
1 P R O C E E D I N G S
2 THE COURT: Good morning, counsel.
3 MR. OSSOLA: Good morning, Your Honor.
4 MR. PAGE: Good morning, Your Honor.
5 THE COURT: Let's see, I think when we recessed,
6 Mr. Page, you were in the midst of your cross examination of
7 Dr. Ford. Do we need to bring him back in?
8 MR. PAGE: Briefly, Your Honor.
9 THE COURT: All right.
10 MR. PAGE: I have a few additional questions.
11 THE COURT: All right, Dr. Ford, if you'd come back up
12 to the witness stand? You're still under affirmation from your
13 testimony on Monday, sir.
14 GARY T. FORD, PH.D., PLAINTIFF'S WITNESS,
15 PREVIOUSLY AFFIRMED, RESUMED
16 CROSS EXAMINATION (Cont'd.)
17 BY MR. PAGE:
18 Q. Good morning, Dr. Ford.
19 A. Good morning, Mr. Page.
20 Q. I'm sorry to have to bring you back, but I just have a few
21 more questions for you. In your direct, you discussed briefly the
22 coding tables that are at tab I of your report. Could I ask you
23 in a little more detail how you go about selecting the categories
24 that you use for coding?
25 A. In general, the coders from in this case Target Research
272
1 Group review perhaps 10 percent of the responses as they come in
2 and read them independently and assign, assign their responses to
3 various categories. At that point, I review what they've done. I
4 look at the categories that they've put together and, if
5 necessary, suggest some changes.
6 Q. Okay. So the categories are a function of what the actual
7 responses are. You don't select them in advance; is that correct?
8 A. That's correct.
9 Q. Okay. And if you'd look at page 5 of tab I of your report?
10 This is Plaintiff's Exhibit 231.
11 A. Yes, I'm there.
12 Q. If you see near the bottom, there's a category 34, and that's
13 labeled "It was a GEICO search/under search/results of search"?
14 A. Yes.
15 Q. Is that the code that you would use if someone indicated that
16 their reason for giving a particular answer was that "GEICO" was
17 the search term?
18 A. Yes. That seems to be -- there may be other codes, too, for
19 it, but that would be one code.
20 Q. But that's what that code would indicate?
21 A. Yes.
22 Q. And when you put your tables together, if, if there are no
23 responses in a given code, do you simply leave that, that code out
24 of the table?
25 A. Or, or put it in as a zero.
273
1 Q. Okay. For example, if you could turn to your Revised Table
2 10, which is at tab 10?
3 A. Yes.
4 Q. These are the responses to Question 1b, which is the reasons
5 why respondents would click on sponsored links --
6 A. Yes.
7 Q. -- if they wanted to purchase automobile insurance?
8 A. Yes.
9 Q. And are there any responses on this table that correspond
10 with the code 35 that we just -- sorry -- 34 that we just saw?
11 A. I would -- let me be more specific. I would assume that code
12 34 would be collapsed into perhaps code -- the first response on
13 Table 1b.
14 Q. That says GEICO --
15 A. In other words, sometimes there's -- if we look on page 5
16 here, there are -- and we look on -- and we look under the "GEICO
17 (Net)," there are several responses that go in the, in the
18 GEICO -- that are combined under the "GEICO (Net)." Now, some of
19 those responses may be folded into the responses on Table 10
20 Revised in the first category.
21 Q. So if I'm correct, you --
22 A. And it could also be in the second one, too.
23 Q. So you coded the answer that I searched for GEICO and the
24 answer "it says GEICO" into a single category?
25 A. It could be. I don't know off the top of my head because of
274
1 the way the nets were brought.
2 Q. Okay. And in fact, in Table 10, the single-most prevalent
3 answer is -- or the two most prevalent answers are "Says get
4 quotes" and "Says save money," correct?
5 A. Yes.
6 Q. Okay. Yesterday, you said that your Questions 2 and 3 were
7 more traditional likelihood of confusion questions; is that
8 correct?
9 A. I did.
10 Q. Okay. In traditional confusion surveys, do you typically
11 present the junior and senior marks sitting next to each other
12 when asking questions?
13 A. No.
14 Q. Okay. But in this case, you really didn't have much choice,
15 did you?
16 A. That is correct.
17 Q. So you end up doing what you call a traditional confusion
18 survey with the allegedly infringing material and the senior mark
19 sitting right next to each other, right?
20 A. They -- there's no other way to do it in this case, and
21 that's also why I thought Question 1 was more than, too. In
22 Question 1, obviously, the respondent has the ability to respond
23 by selecting the place they would click first from the organic
24 results or from the sponsored links, and so I think you have to
25 have -- that's one of the reasons I used both of those questions.
275
1 Q. I see. So you feel it's less of a problem for Question 1
2 than for Questions 2 or 3?
3 A. I don't necessarily feel it's a problem for Questions 2 and
4 3. I think it was important to do that, and it's also important
5 to have a control to try to get at those kinds of issues.
6 Q. But you would, of course, expect that you'd get higher
7 affiliation answers when people are staring at the name than an
8 unaided affiliation question; would you agree?
9 A. I would think that in general, that's so, but given the
10 characteristics of this situation, which is that they entered
11 "GEICO" as a search term and they were asked for one of the
12 sponsored links where they would expect to go if they clicked on
13 that sponsored link, I don't see any practical way how to do that
14 other than what was done.
15 Q. I see. Could I get a look at your survey questionnaire,
16 which is at tab G, I believe? Tab G. I want to look at your
17 Question 3a.
18 A. Excuse me, 2a?
19 Q. No, 3a is actually the one I have.
20 A. That's it.
21 Q. That's perfect.
22 Let me see if I understand this correctly. Question 3
23 was only asked of people who had not already answered "GEICO" to
24 Question 1 or 2, correct?
25 A. To Question 2 only.
276
1 Q. Question 2. So any respondent who answered "GEICO" to
2 Question 2, they were done. They never got to Question 3?
3 A. They were, they were asked 2b and 2c, but other than that,
4 they were done.
5 Q. Right. So that the only people who were answering Question 3
6 were people who got what I'll call the right answer being asked
7 about the sponsored links, in other words, gave an answer that
8 wasn't "GEICO" but for the most part was the actual link that
9 those ads went to, correct?
10 A. The only people who were asked Question 3 were people who had
11 not given a "GEICO" response to Question 2, which is what it says
12 in the directions to that question.
13 Q. Okay. So Question 3a was asked of people who, for example,
14 had said "netquote," and you then asked them, "Do you think the
15 company that sponsors this listing is associated or connected with
16 any other company or companies?" Correct?
17 A. That's correct.
18 Q. Aren't you just telling those respondents, "No, wrong answer.
19 Try again"?
20 A. Well, you can -- you have the opportunity to look at the
21 results individually for those, those responses, and there is --
22 it seems to me those are traditional questions that we would get
23 in a likelihood of confusion survey, generally followed by a third
24 question that asked whether there's a need, permission kind of
25 aspects.
277
1 So in this case, I was asking two questions that are
2 frequently asked in these types of studies, and the results are
3 reported separately. It can be aggregated separately so we can
4 determine whether there is any large jump in responses for the
5 people who were asked the second question and not the first.
6 Q. But in fact, you've only asked this -- you've turned
7 everybody who got the right answer and said, "Try again," and then
8 you've added them in as confused to the people who said "GEICO" to
9 the second, while at the same time, you haven't asked this
10 question of people who said "GEICO," so you haven't given them an
11 opportunity to say, "Well, yes, I think it may also be netquote."
12 Correct?
13 A. That's correct.
14 MR. PAGE: Okay. I have no further questions.
15 THE WITNESS: But -- thank you.
16 THE COURT: All right. Mr. Ossola, any redirect?
17 MR. OSSOLA: No, Your Honor.
18 THE COURT: All right, thank you, Dr. Ford. You may
19 step down.
20 THE WITNESS: You're welcome. Thank you.
21 (Witness excused.)
22 THE COURT: All right, Mr. Ossola, is there any other
23 evidence GEICO wants to present in its case-in-chief?
24 MR. OSSOLA: No, Your Honor, other than the evidence
25 that's already been submitted on the papers beyond that which is
278
1 presented in open court.
2 THE COURT: That's fine. And just for the record,
3 because this case went in in a somewhat -- the plaintiff's case
4 has gone in in a somewhat unusual fashion, for any appellate
5 purposes if that were to be necessary, we need to clear up what
6 exhibits are actually in evidence.
7 In particular, Mr. Page, you referred to several of your
8 exhibits during the course of the plaintiff's case. I believe I
9 have every one of those recorded, but we would just need to make
10 sure that that's done, and you should check with Ms. Travers if
11 there's any question about that. All right?
12 MR. OSSOLA: Yes, Your Honor.
13 THE COURT: All right. Mr. Page?
14 MR. PAGE: Your Honor, at this time, Google would like
15 to move for judgment as a matter of law in this case.
16 THE COURT: All right, sir.
17 MR. PAGE: At the opening of this case, I said that the
18 Court would be able to dispose of this case before hearing
19 testimony on damages. That got a bit of a laugh, but I was quite
20 serious. Now that the Court has heard plaintiff's case, we
21 believe you can rule in Google's favor as a matter of law.
22 On the key issue in this case, whether the use of
23 trademarks as keywords violates the Lanham Act, GEICO's own
24 evidence proves Google's case. To the extent users are confused
25 at all, that confusion comes from either the use of GEICO's
279
1 trademark in the text of the sponsored link or from the natural
2 and nonactionable assumption that any site that offers to compare
3 insurance rates will likely include GEICO.
4 It's important to note this assumption is only wrong
5 because GEICO as a business decision has chosen to make it wrong.
6 That's their choice. As you heard from their witnesses, it's an
7 integral part of their business strategy to sell only direct to
8 customers and not to allow other companies to quote rates that may
9 not be accurate.
10 It's a perfectly legitimate strategy and, combined with
11 a great marketing campaign, has resulted in a remarkable success
12 story. That's their choice, but Google cannot be liable for
13 confusion that's a natural result of that business strategy.
14 What GEICO's evidence shows is that the use of
15 trademarks as keywords does not itself create confusion. As they
16 bear the burden of proof on this point, that's the end of the
17 analysis.
18 If the Court were to deny this motion, this afternoon or
19 tomorrow you would hear from Google's survey expert, Dr. Jacoby,
20 whose own survey clearly establishes the same point. When
21 respondents enter the word "GEICO" as a search term and are
22 presented with sponsored links that offer car insurance quotes,
23 some of them say yes when asked if they think they can get
24 information about GEICO there, but they also say yes at a
25 frequency that is actually slightly higher when asked the same
280
1 question about Allstate, even though "Allstate" appears nowhere on
2 the search term, the organic results, or on the sponsored links.
3 So on the question of whether the use of trademarks as
4 keywords violates the Lanham Act, the answer is clearly no.
5 Turning to the question of GEICO's mark in the text of
6 ads and whether that gives rise to confusion, we believe that the
7 flaws in GEICO's survey, which is their only evidence on this
8 point, mandate a finding as a matter of law in Google's favor, but
9 the Court need not throw out that evidence in order to rule in our
10 favor. As GEICO concedes, such use is already banned by Google's
11 trademark policy, and Google vigorously enforces that policy.
12 GEICO has told you that nonetheless, some ads
13 occasionally slip through, although they have presented no
14 evidence of how often that occurs, but as I noted in opening
15 argument, the inability to achieve perfect enforcement of that
16 policy does not give rise to contributory liability. In order to
17 prevail on that claim, GEICO would have to establish far more:
18 that Google affirmatively encouraged or knowingly assisted in a
19 violation of trademark law by the alleged infringers.
20 There's no such evidence before the Court because Google
21 does not condone or encourage infringement. Quite to the
22 contrary, we were the first search engine to implement a trademark
23 enforcement policy, both because it is the right thing to do and
24 because it makes business sense.
25 Trademark owners are our own customers on the
281
1 advertising side, just as the users are our customers on the
2 search side. Thus, we carefully balance the dual goals of
3 protecting trademark rights on the one hand with providing our
4 users with the most complete and relevant information possible on
5 the other hand.
6 We believe that our current trademark policy strikes
7 exactly the correct balance, neither over- nor under-protecting
8 trademark rights. Therefore, we ask that the Court enter judgment
9 as a matter of law in Google's favor. Thank you.
10 THE COURT: All right, thank you.
11 Mr. Ossola?
12 MR. OSSOLA: Your Honor, good morning. Mr. Page starts
13 out by describing the issue that he believes there's a failure of
14 proof with respect to as a matter of law as whether the use of
15 keywords -- whether the use of trademarks as keywords in and of
16 itself gives rise to a likelihood of confusion and claims that
17 GEICO has failed to establish that proposition, but I would submit
18 to you that that is not the proposition that GEICO is advancing in
19 this case.
20 The claim of infringement, the claim of likelihood of
21 confusion is one that is based on the sponsored links that result
22 from under Google's system the user's entry of the search term
23 "GEICO," and it is the relationship between the search term and
24 those sponsored listings that gives rise to the likelihood of
25 confusion.
282
1 One cannot in our view view -- look at the question of
2 the use of "GEICO" as a keyword by itself as presenting the real
3 question to be decided based on the evidence presented in this
4 case.
5 And as we have attempted to make clear, it is the two
6 categories of sponsored listings that at least thus far are
7 generated by GEICO as a keyword that gives rise to the likelihood
8 of confusion. One of those categories is the bulk of the
9 sponsored listings that the searches that we have put into
10 evidence demonstrates, and that is those that include "GEICO" in
11 the text, the title or heading of the sponsored listing.
12 As to that category of sponsored listings, as we said in
13 our opening statement, it seems to me that there is a complete
14 sufficiency of proof, particularly when there is no proof to the
15 contrary that will be offered by Google to conclude that there is
16 a likelihood of confusion, and I say that for two reasons.
17 One is simply by looking at the searches themselves that
18 had -- that establish a connection between a search term and a
19 sponsored listing. Secondly, Dr. Ford's survey, if it proves
20 nothing else, certainly establishes based on the sponsored
21 listings that he tested, both that included the "GEICO" as a
22 search term and the one that did not, that there is an
23 overwhelming likelihood of confusion that is associated with
24 sponsored listings that include the trademark in the text.
25 That is something that's established by his survey.
283
1 It's established by the other evidence submitted, the searches
2 themselves, and particularly given the evidence that came in with
3 respect to GEICO's focal point of its advertising of its business
4 on GEICO as a, a trademark that is used to prompt potential
5 customers to search for a rate quote from GEICO, the fact that a
6 GEICO rate quote cannot be found, the evidence shows, from any
7 site other than GEICO's, and the fact that a consumer would be
8 misled into believing that a third-party source that is offering a
9 comparison of rate quotes that either refers to GEICO specifically
10 or does not refer to GEICO specifically suggests to the consumer
11 that they can find a rate quote from GEICO at those sites, and
12 it's clear, I think, from the evidence that they cannot and that
13 those ads are inherently misleading.
14 Now, Mr. Page refers to the fact that Google already --
15 in its policy already provides that sponsored listings that have
16 the trademark in the text should not be displayed, and as we
17 pointed out, that has always been the case under Google's policy.
18 Its policy did not change with respect to that.
19 And I think the evidence is clear that that has
20 happened. It has continued to happen, and despite the policy, the
21 reality is that those sponsored listings have been displayed
22 despite the objection of GEICO, and in evidence are those
23 objections, in evidence are the numerous instances in which
24 sponsored listings containing the search term "GEICO" continued to
25 appear.
284
1 So far from slipping through, it seems to me that the
2 bulk of the search terms that are in evidence establish that those
3 are the type of categories of listings, that is, with "GEICO" in
4 the title, that have, in fact, occurred in the marketplace and, we
5 believe, clearly give rise to a likelihood of confusion.
6 With respect to the question of whether Dr. Ford's
7 survey establishes a likelihood of confusion as to those sponsored
8 listings that do not have "GEICO" in the title, we submit that
9 Dr. Ford's survey is sufficient to prove that proposition. He
10 did, in fact, test both sponsored listings that include and do not
11 include "GEICO" in the text. The result -- the respondents in the
12 survey were asked independently to react to each of those five
13 sponsored listings.
14 The evidence shows that the responses with respect to a
15 source affiliation were not appreciably different among those
16 respondents. There was a slight dimunition in the level of
17 confusion associated with the fifth sponsored link, but it's still
18 in excess of 50 percent.
19 So we believe that the evidence clearly establishes that
20 there is a likelihood of confusion based both upon the use of
21 "GEICO" in the text, and in this case, given the fact that "GEICO"
22 is a mark associated with rate quotes, that it is also sufficient
23 to establish a likelihood of confusion with respect to those
24 sponsored listings that do not contain "GEICO" in the text.
25 Thank you, Your Honor.
285
1 THE COURT: All right. Well, since this is a bench
2 trial, I could be wrong, but I'm going to assume that this motion
3 is raised under rule 52(c). I think that's the right rule.
4 MR. PAGE: That is correct, Your Honor.
5 THE COURT: And what the Court is being called upon to
6 do at this point in the trial is to evaluate the strength of the
7 plaintiff's evidence to see whether or not the case should go
8 forward on any or all of the issues in this case, and having
9 carefully considered the plaintiff's case, the Court is going to
10 grant in part and deny in part the defendant's motion.
11 As we all know, there are five essential elements to a
12 Lanham Act claim. First, the plaintiff must establish that it
13 possesses a mark that is protectable, and of course, that's not an
14 issue in this case. "GEICO" clearly is.
15 The remaining elements require basically a focus on what
16 the defendant does. Question No. -- the second element is that
17 the defendant uses the mark; three, that the defendant's use of
18 the mark in commerce; four, that the defendant's use of the mark
19 is in connection with the sale, offering for sale, distribution,
20 or advertising of goods and services; and fifth and what has in my
21 mind been the key issue in why I could not grant summary judgment
22 initially in this case or grant the motion to dismiss that was
23 previously filed was that the use of the mark is done in a manner
24 that's likely to cause confusion.
25 Having heard the plaintiff's case, the Court is
286
1 satisfied that the plaintiff has not established that the mere use
2 of its trademark by Google as a search word or keyword or even
3 using it in their AdWord program standing alone violates the
4 Lanham Act because that activity in and of itself, there's no
5 evidence that that activity standing alone causes confusion.
6 The Court also finds that there was insufficient
7 evidence presented in the plaintiff's case to let this case go
8 forward on the question of whether Google violated the Lanham Act
9 after it barred -- or after it began to bar the use of the GEICO
10 mark from either the titles or the text of the sponsored ads that
11 appear as a result of use of the AdWord program.
12 And the reason I find insufficient evidence of that, as
13 we sort of -- I sort of indicated during the testimony of
14 Dr. Ford's, is that, frankly, Dr. Ford either wasn't asked or
15 chose not to actually query that particular issue because the
16 survey focused on either ads that had "GEICO" in the title -- I
17 mean, the page that people were looking at had five sponsored ads,
18 the first four of which had "GEICO" either in the title and/or in
19 the text. The fifth one did not, but it was on a page that had
20 the four previous sponsored links.
21 The test -- the control in this case used Nike, but what
22 I don't feel was presented to the Court that needed to be
23 presented to the Court would have been ads of insurance that did
24 not have "GEICO" in it, and I was not satisfied therefore that the
25 plaintiff's evidence was sufficient.
287
1 However, as to the narrow issue of whether the plaintiff
2 has presented sufficient evidence at this point to let the case
3 continue on the issue of whether Google is liable for violating
4 the Lanham Act for those sponsored ads using GEICO's name in
5 either the title or the text that appear next to a -- and a GEICO
6 organic listing as a result of the AdWord program, at this point,
7 there has been enough evidence of confusion as a result of
8 Dr. Ford's report to deny the motion and to allow the case to
9 continue so the Court can consider the defendant's evidence as to
10 whether or not this situation creates a likelihood of confusion in
11 the marketplace and, if so, what damages might be appropriate.
12 Now, counsel, is my ruling clear, and any questions
13 about what I have just done?
14 MR. OSSOLA: Your Honor, I have just one.
15 THE COURT: Yes.
16 MR. OSSOLA: You made reference to the -- what you
17 believed to be the insufficiency of evidence after Google began to
18 bar the use of the trademark in the sponsored link. I think
19 that's what you said. I just want to make clear that Google has
20 always barred --
21 THE COURT: And I should clarify that. I realize that,
22 but in this particular case, they -- I meant to confine that to
23 the facts of this case; that is, I didn't mean to suggest that
24 Google did not have that policy in place, but the reality was that
25 we do have evidence that there were -- "GEICO" was being used in
288
1 titles and in text until Google was advised to stop that, and then
2 following its own policy, it implemented that policy.
3 MR. OSSOLA: I understand.
4 THE COURT: All right.
5 MR. OSSOLA: But the insufficiency evidence on that
6 point is that with respect to those sponsored listings that do not
7 have "GEICO" in the text or heading of the ad, as to those -- that
8 category of sponsored listings, the Court is finding that there is
9 not sufficient evidence of likelihood of confusion.
10 THE COURT: In this case.
11 MR. OSSOLA: In this case.
12 THE COURT: Correct.
13 MR. OSSOLA: Thank you.
14 THE COURT: All right?
15 Mr. Page, was there anything you needed clarified?
16 MR. PAGE: No, I think that covers it, Your Honor.
17 THE COURT: All right. Now, I think as we all know,
18 there's serious legal issues that are involved in this decision
19 that I've just rendered, in particular, the status of how
20 trademarks as used in keywords, etc., and I think I would like
21 time to write a more detailed opinion on these legal issues, and
22 so depending upon how you-all want to proceed, as you know, I've
23 also been encouraging you to see if you can resolve the case, and
24 I've given you now some very clear parameters as to how the case
25 would continue.
289
1 Unless there's any objection, what I propose is that we
2 terminate the trial at this point, I don't mean end it, but stop
3 right now what we're doing to give the Court a brief amount of
4 time, which given the holiday season might be two or three weeks,
5 although we're going to try to do it sooner than that, to get a
6 written opinion out on this ruling, consistent with this ruling,
7 and to allow you-all the opportunity to see whether or not there
8 can be a resolution of what is left in the case.
9 Does that meet with all of your approval?
10 MR. PAGE: Yes, Your Honor, that would be fine.
11 MR. OSSOLA: Yes, Your Honor.
12 THE COURT: All right. Now, I'm not going to set
13 another date at this point for when we would start up Round 2 if
14 that becomes necessary. We'll get the opinion out as quickly as
15 possible, and then you can get back to me as to how you want to
16 proceed.
17 In the meantime, since this courtroom is not going to be
18 used again until January 3, and I'm hoping we have it resolved by
19 then, you can if you want leave your exhibits up here. If the
20 case ultimately has been resolved, then there's no need for any of
21 the exhibits to stay at the courthouse, and we'll ask you to come
22 and get them.
23 If the case goes on to the, what I would call the second
24 phase, then obviously, we'll need to get the appellate record
25 cleaned up in terms of just what exhibits are in or out. All
290
1 right?
2 MR. OSSOLA: Yes.
3 THE COURT: Anything further we need to address at this
4 point?
5 MR. OSSOLA: Your Honor, just the second phase, if we
6 get there, would be -- would include damages as well; is that
7 correct?
8 THE COURT: Yes. It would be the defendant's
9 opportunity to rebut liability, and obviously, if I were to find
10 that as a matter of law there was no issue left on liability and
11 find in the defendant's favor, then we wouldn't have to have a
12 damages phase, but if I find liability, then there would be a
13 damages phase. All right?
14 MR. OSSOLA: What's been established to date then is
15 that the category of sponsored listings that have "GEICO" in the
16 title and text are -- have been found to be confusing and
17 misleading under the Lanham Act?
18 THE COURT: No. I have found that you've presented
19 enough evidence at this point to avoid a motion for judgment as a
20 matter of law. In other words, there certainly is a prima facie
21 case that you've established that they are confusing based on
22 Dr. Ford's survey.
23 Now, I haven't heard the rebuttal evidence, so I'm not
24 making a specific finding on that. I'm at this point weighing the
25 quality of the evidence, the amount of evidence. You've presented
291
1 enough to let the case go forward.
2 MR. OSSOLA: Thank you, Your Honor.
3 THE COURT: All right?
4 MR. PAGE: Thank you, Your Honor.
5 THE COURT: All right? And, counsel, that one piece of
6 paper that was on each of your chairs was a joke. It was not
7 serious. But we happened to see it and thought it was
8 appropriate.
9 All right, I wish you-all a happy holiday, and we'll get
10 back to you as quickly as possible.
11 We'll recess court for the day.
12 MR. OSSOLA: Thank you, Your Honor.
13 MR. PAGE: Thank you, Your Honor.
14 (Recess from 10:31 a.m., until 11:09 a.m.)
15 THE COURT: All right, for the record, Mr. Ossola has
16 brought to the Court's attention the following matter: My
17 understanding, Mr. Ossola, is that your view of the defendant's
18 evidence is that Google will not be able to offer any evidence to
19 counter GEICO's evidence that those sponsored sites that contain
20 "GEICO" either in their title or in their text that have been at
21 issue in this case are confusing. Is that correct?
22 MR. OSSOLA: That's correct, Your Honor.
23 THE COURT: And it's my understanding that, what, the
24 expert -- is it Dr. Jacoby, Mr. Page? Whose report would have --
25 MR. PAGE: That's correct. It would be Dr. Jacoby's
292
1 report. Mr. Ossola is correct. Our survey addressed Google's
2 current policy, so the sponsored links that we tested do not
3 contain the word "GEICO." We were testing the keyword issue.
4 THE COURT: All right.
5 MR. PAGE: So Dr. Jacoby's survey says nothing either
6 way as to whether a link with GEICO is confusing.
7 THE COURT: All right. Then the Court should have
8 clarified its earlier -- well, it couldn't clarify -- should have
9 said in its earlier ruling and says now in response to your motion
10 that the evidence before this Court does establish that those
11 sponsored sites that contain "GEICO" either in the title or in the
12 text are likely to confuse for purposes of the Lanham Act
13 requirements, and therefore, to the extent that the defendant's
14 motion was based on that issue, that will also be denied.
15 And that reduces the issues that would have to go
16 forward were there -- if there is a second phase to this trial.
17 The sole issues that would remain -- and correct me if I'm
18 wrong -- are, No. 1, whether or not Google is liable for any
19 Lanham Act violation based upon those sponsored sites, and 2, if
20 Google were liable, then what damages would be appropriate.
21 Now, is that a correct phrasing from your standpoints as
22 to what would be left in this case?
23 MR. PAGE: That is correct, Your Honor. The issue that
24 will remain is, is since the Court has found a likelihood of
25 confusion for ads that have "GEICO" in them, the question is
293
1 whether Google is contributorily liable or whether only the
2 advertiser would be liable.
3 THE COURT: Well, actually, because the advertisers are
4 not in this case, the only question before us would be whether or
5 not Google is liable.
6 MR. PAGE: That's correct, Your Honor.
7 THE COURT: Right. Mr. Ossola?
8 MR. OSSOLA: I believe that's right. Under contributory
9 liability or any other theory that may be asserted by GEICO, that
10 remains to be seen.
11 THE COURT: Correct.
12 MR. OSSOLA: Phase 2.
13 THE COURT: So there has been -- just so we're clear for
14 the record, there has been no finding of liability -- there's been
15 no finding that Google is liable at this point. What we have
16 found, however, is that this particular group of sponsored sites
17 does violate the Lanham Act.
18 MR. OSSOLA: Yes, Your Honor.
19 THE COURT: All right?
20 MR. PAGE: That's correct.
21 THE COURT: Do you think we need to clarify anything
22 further for the record?
23 MR. OSSOLA: No, Your Honor, thank you.
24 THE COURT: All right. Thank you very much for bringing
25 that to our attention. We'll recess court once again.
294
1 (Recess at 11:12 a.m.)
2
3 CERTIFICATE OF THE REPORTER
4 I certify that the foregoing is a correct transcript of the
5 record of proceedings in the above-entitled matter.
6
7
8
Anneliese J. Thomson
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
295
1 I N D E X
2 DIRECT CROSS REDIRECT RECROSS
3 WITNESS ON BEHALF OF
THE PLAINTIFF:
4
Gary T. Ford, Ph.D. 271
5 (Resumed)
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
269
UNITED STATES DISTRICT COURT
FOR THE EASTERN DISTRICT OF VIRGINIA
ALEXANDRIA DIVISION
GOVERNMENT EMPLOYEES . Civil Action No. 1:04cv507
INSURANCE COMPANY, .
.
Plaintiff, .
.
vs. . Alexandria, Virginia
. December 15, 2004
GOOGLE, INC., . 10:00 a.m.
.
Defendant. .
.
. . . . . . . . . . .
TRANSCRIPT OF BENCH TRIAL
BEFORE THE HONORABLE LEONIE M. BRINKEMA
UNITED STATES DISTRICT JUDGE
VOLUME II
APPEARANCES:
FOR THE PLAINTIFF: CHARLES D. OSSOLA, ESQ.
CHRISTOPHER WINTERS, ESQ.
MICHAEL J. ALLAN, ESQ.
ROBERTA L. HORTON, ESQ.
Arnold & Porter LLP
555 Twelfth Street, N.W.
Washington, D.C. 20004
and
JOHN F. ANDERSON, ESQ.
Troutman Sanders LLP
1660 International Drive, Suite 600
McLean, VA 22102
and
JONATHAN L. SHAFNER, ESQ.
Government Employees Companies
One Geico Plaza
Washington, D.C. 20076
(APPEARANCES CONT'D. ON FOLLOWING PAGE)
(Pages 269 - 295)
COMPUTERIZED TRANSCRIPTION OF STENOGRAPHIC NOTES
270
1 FOR THE DEFENDANT: MICHAEL H. PAGE, ESQ.
JOSHUA H. LERNER, ESQ.
2 ANJALI S. SAKARIA, ESQ.
Keker & Van Nest, LLP
3 710 Sansome Street
San Francisco, CA 94111-1704
4 and
MELANIE D. COATES, ESQ.
5 Wilmer Cutler Pickering Hale and
Dorr LLP
6 1600 Tysons Boulevard, Suite 1000
McLean, VA 22102
7 and
MICHAEL S. KWUN, ESQ.
8 Google
1600 Amphitheatre
9 Mountain View, CA 94043
10
ALSO PRESENT: RHONDA L. ANDREW
11 DAVID GORDON
12
OFFICIAL COURT REPORTER: ANNELIESE J. THOMSON, RDR, CRR
13 U.S. District Court, Fifth Floor
401 Courthouse Square
14 Alexandria, VA 22314
(703)299-8595
15
16
17
18
19
20
21
22
23
24
25
271
1 P R O C E E D I N G S
2 THE COURT: Good morning, counsel.
3 MR. OSSOLA: Good morning, Your Honor.
4 MR. PAGE: Good morning, Your Honor.
5 THE COURT: Let's see, I think when we recessed,
6 Mr. Page, you were in the midst of your cross examination of
7 Dr. Ford. Do we need to bring him back in?
8 MR. PAGE: Briefly, Your Honor.
9 THE COURT: All right.
10 MR. PAGE: I have a few additional questions.
11 THE COURT: All right, Dr. Ford, if you'd come back up
12 to the witness stand? You're still under affirmation from your
13 testimony on Monday, sir.
14 GARY T. FORD, PH.D., PLAINTIFF'S WITNESS,
15 PREVIOUSLY AFFIRMED, RESUMED
16 CROSS EXAMINATION (Cont'd.)
17 BY MR. PAGE:
18 Q. Good morning, Dr. Ford.
19 A. Good morning, Mr. Page.
20 Q. I'm sorry to have to bring you back, but I just have a few
21 more questions for you. In your direct, you discussed briefly the
22 coding tables that are at tab I of your report. Could I ask you
23 in a little more detail how you go about selecting the categories
24 that you use for coding?
25 A. In general, the coders from in this case Target Research
272
1 Group review perhaps 10 percent of the responses as they come in
2 and read them independently and assign, assign their responses to
3 various categories. At that point, I review what they've done. I
4 look at the categories that they've put together and, if
5 necessary, suggest some changes.
6 Q. Okay. So the categories are a function of what the actual
7 responses are. You don't select them in advance; is that correct?
8 A. That's correct.
9 Q. Okay. And if you'd look at page 5 of tab I of your report?
10 This is Plaintiff's Exhibit 231.
11 A. Yes, I'm there.
12 Q. If you see near the bottom, there's a category 34, and that's
13 labeled "It was a GEICO search/under search/results of search"?
14 A. Yes.
15 Q. Is that the code that you would use if someone indicated that
16 their reason for giving a particular answer was that "GEICO" was
17 the search term?
18 A. Yes. That seems to be -- there may be other codes, too, for
19 it, but that would be one code.
20 Q. But that's what that code would indicate?
21 A. Yes.
22 Q. And when you put your tables together, if, if there are no
23 responses in a given code, do you simply leave that, that code out
24 of the table?
25 A. Or, or put it in as a zero.
273
1 Q. Okay. For example, if you could turn to your Revised Table
2 10, which is at tab 10?
3 A. Yes.
4 Q. These are the responses to Question 1b, which is the reasons
5 why respondents would click on sponsored links --
6 A. Yes.
7 Q. -- if they wanted to purchase automobile insurance?
8 A. Yes.
9 Q. And are there any responses on this table that correspond
10 with the code 35 that we just -- sorry -- 34 that we just saw?
11 A. I would -- let me be more specific. I would assume that code
12 34 would be collapsed into perhaps code -- the first response on
13 Table 1b.
14 Q. That says GEICO --
15 A. In other words, sometimes there's -- if we look on page 5
16 here, there are -- and we look on -- and we look under the "GEICO
17 (Net)," there are several responses that go in the, in the
18 GEICO -- that are combined under the "GEICO (Net)." Now, some of
19 those responses may be folded into the responses on Table 10
20 Revised in the first category.
21 Q. So if I'm correct, you --
22 A. And it could also be in the second one, too.
23 Q. So you coded the answer that I searched for GEICO and the
24 answer "it says GEICO" into a single category?
25 A. It could be. I don't know off the top of my head because of
274
1 the way the nets were brought.
2 Q. Okay. And in fact, in Table 10, the single-most prevalent
3 answer is -- or the two most prevalent answers are "Says get
4 quotes" and "Says save money," correct?
5 A. Yes.
6 Q. Okay. Yesterday, you said that your Questions 2 and 3 were
7 more traditional likelihood of confusion questions; is that
8 correct?
9 A. I did.
10 Q. Okay. In traditional confusion surveys, do you typically
11 present the junior and senior marks sitting next to each other
12 when asking questions?
13 A. No.
14 Q. Okay. But in this case, you really didn't have much choice,
15 did you?
16 A. That is correct.
17 Q. So you end up doing what you call a traditional confusion
18 survey with the allegedly infringing material and the senior mark
19 sitting right next to each other, right?
20 A. They -- there's no other way to do it in this case, and
21 that's also why I thought Question 1 was more than, too. In
22 Question 1, obviously, the respondent has the ability to respond
23 by selecting the place they would click first from the organic
24 results or from the sponsored links, and so I think you have to
25 have -- that's one of the reasons I used both of those questions.
275
1 Q. I see. So you feel it's less of a problem for Question 1
2 than for Questions 2 or 3?
3 A. I don't necessarily feel it's a problem for Questions 2 and
4 3. I think it was important to do that, and it's also important
5 to have a control to try to get at those kinds of issues.
6 Q. But you would, of course, expect that you'd get higher
7 affiliation answers when people are staring at the name than an
8 unaided affiliation question; would you agree?
9 A. I would think that in general, that's so, but given the
10 characteristics of this situation, which is that they entered
11 "GEICO" as a search term and they were asked for one of the
12 sponsored links where they would expect to go if they clicked on
13 that sponsored link, I don't see any practical way how to do that
14 other than what was done.
15 Q. I see. Could I get a look at your survey questionnaire,
16 which is at tab G, I believe? Tab G. I want to look at your
17 Question 3a.
18 A. Excuse me, 2a?
19 Q. No, 3a is actually the one I have.
20 A. That's it.
21 Q. That's perfect.
22 Let me see if I understand this correctly. Question 3
23 was only asked of people who had not already answered "GEICO" to
24 Question 1 or 2, correct?
25 A. To Question 2 only.
276
1 Q. Question 2. So any respondent who answered "GEICO" to
2 Question 2, they were done. They never got to Question 3?
3 A. They were, they were asked 2b and 2c, but other than that,
4 they were done.
5 Q. Right. So that the only people who were answering Question 3
6 were people who got what I'll call the right answer being asked
7 about the sponsored links, in other words, gave an answer that
8 wasn't "GEICO" but for the most part was the actual link that
9 those ads went to, correct?
10 A. The only people who were asked Question 3 were people who had
11 not given a "GEICO" response to Question 2, which is what it says
12 in the directions to that question.
13 Q. Okay. So Question 3a was asked of people who, for example,
14 had said "netquote," and you then asked them, "Do you think the
15 company that sponsors this listing is associated or connected with
16 any other company or companies?" Correct?
17 A. That's correct.
18 Q. Aren't you just telling those respondents, "No, wrong answer.
19 Try again"?
20 A. Well, you can -- you have the opportunity to look at the
21 results individually for those, those responses, and there is --
22 it seems to me those are traditional questions that we would get
23 in a likelihood of confusion survey, generally followed by a third
24 question that asked whether there's a need, permission kind of
25 aspects.
277
1 So in this case, I was asking two questions that are
2 frequently asked in these types of studies, and the results are
3 reported separately. It can be aggregated separately so we can
4 determine whether there is any large jump in responses for the
5 people who were asked the second question and not the first.
6 Q. But in fact, you've only asked this -- you've turned
7 everybody who got the right answer and said, "Try again," and then
8 you've added them in as confused to the people who said "GEICO" to
9 the second, while at the same time, you haven't asked this
10 question of people who said "GEICO," so you haven't given them an
11 opportunity to say, "Well, yes, I think it may also be netquote."
12 Correct?
13 A. That's correct.
14 MR. PAGE: Okay. I have no further questions.
15 THE WITNESS: But -- thank you.
16 THE COURT: All right. Mr. Ossola, any redirect?
17 MR. OSSOLA: No, Your Honor.
18 THE COURT: All right, thank you, Dr. Ford. You may
19 step down.
20 THE WITNESS: You're welcome. Thank you.
21 (Witness excused.)
22 THE COURT: All right, Mr. Ossola, is there any other
23 evidence GEICO wants to present in its case-in-chief?
24 MR. OSSOLA: No, Your Honor, other than the evidence
25 that's already been submitted on the papers beyond that which is
278
1 presented in open court.
2 THE COURT: That's fine. And just for the record,
3 because this case went in in a somewhat -- the plaintiff's case
4 has gone in in a somewhat unusual fashion, for any appellate
5 purposes if that were to be necessary, we need to clear up what
6 exhibits are actually in evidence.
7 In particular, Mr. Page, you referred to several of your
8 exhibits during the course of the plaintiff's case. I believe I
9 have every one of those recorded, but we would just need to make
10 sure that that's done, and you should check with Ms. Travers if
11 there's any question about that. All right?
12 MR. OSSOLA: Yes, Your Honor.
13 THE COURT: All right. Mr. Page?
14 MR. PAGE: Your Honor, at this time, Google would like
15 to move for judgment as a matter of law in this case.
16 THE COURT: All right, sir.
17 MR. PAGE: At the opening of this case, I said that the
18 Court would be able to dispose of this case before hearing
19 testimony on damages. That got a bit of a laugh, but I was quite
20 serious. Now that the Court has heard plaintiff's case, we
21 believe you can rule in Google's favor as a matter of law.
22 On the key issue in this case, whether the use of
23 trademarks as keywords violates the Lanham Act, GEICO's own
24 evidence proves Google's case. To the extent users are confused
25 at all, that confusion comes from either the use of GEICO's
279
1 trademark in the text of the sponsored link or from the natural
2 and nonactionable assumption that any site that offers to compare
3 insurance rates will likely include GEICO.
4 It's important to note this assumption is only wrong
5 because GEICO as a business decision has chosen to make it wrong.
6 That's their choice. As you heard from their witnesses, it's an
7 integral part of their business strategy to sell only direct to
8 customers and not to allow other companies to quote rates that may
9 not be accurate.
10 It's a perfectly legitimate strategy and, combined with
11 a great marketing campaign, has resulted in a remarkable success
12 story. That's their choice, but Google cannot be liable for
13 confusion that's a natural result of that business strategy.
14 What GEICO's evidence shows is that the use of
15 trademarks as keywords does not itself create confusion. As they
16 bear the burden of proof on this point, that's the end of the
17 analysis.
18 If the Court were to deny this motion, this afternoon or
19 tomorrow you would hear from Google's survey expert, Dr. Jacoby,
20 whose own survey clearly establishes the same point. When
21 respondents enter the word "GEICO" as a search term and are
22 presented with sponsored links that offer car insurance quotes,
23 some of them say yes when asked if they think they can get
24 information about GEICO there, but they also say yes at a
25 frequency that is actually slightly higher when asked the same
280
1 question about Allstate, even though "Allstate" appears nowhere on
2 the search term, the organic results, or on the sponsored links.
3 So on the question of whether the use of trademarks as
4 keywords violates the Lanham Act, the answer is clearly no.
5 Turning to the question of GEICO's mark in the text of
6 ads and whether that gives rise to confusion, we believe that the
7 flaws in GEICO's survey, which is their only evidence on this
8 point, mandate a finding as a matter of law in Google's favor, but
9 the Court need not throw out that evidence in order to rule in our
10 favor. As GEICO concedes, such use is already banned by Google's
11 trademark policy, and Google vigorously enforces that policy.
12 GEICO has told you that nonetheless, some ads
13 occasionally slip through, although they have presented no
14 evidence of how often that occurs, but as I noted in opening
15 argument, the inability to achieve perfect enforcement of that
16 policy does not give rise to contributory liability. In order to
17 prevail on that claim, GEICO would have to establish far more:
18 that Google affirmatively encouraged or knowingly assisted in a
19 violation of trademark law by the alleged infringers.
20 There's no such evidence before the Court because Google
21 does not condone or encourage infringement. Quite to the
22 contrary, we were the first search engine to implement a trademark
23 enforcement policy, both because it is the right thing to do and
24 because it makes business sense.
25 Trademark owners are our own customers on the
281
1 advertising side, just as the users are our customers on the
2 search side. Thus, we carefully balance the dual goals of
3 protecting trademark rights on the one hand with providing our
4 users with the most complete and relevant information possible on
5 the other hand.
6 We believe that our current trademark policy strikes
7 exactly the correct balance, neither over- nor under-protecting
8 trademark rights. Therefore, we ask that the Court enter judgment
9 as a matter of law in Google's favor. Thank you.
10 THE COURT: All right, thank you.
11 Mr. Ossola?
12 MR. OSSOLA: Your Honor, good morning. Mr. Page starts
13 out by describing the issue that he believes there's a failure of
14 proof with respect to as a matter of law as whether the use of
15 keywords -- whether the use of trademarks as keywords in and of
16 itself gives rise to a likelihood of confusion and claims that
17 GEICO has failed to establish that proposition, but I would submit
18 to you that that is not the proposition that GEICO is advancing in
19 this case.
20 The claim of infringement, the claim of likelihood of
21 confusion is one that is based on the sponsored links that result
22 from under Google's system the user's entry of the search term
23 "GEICO," and it is the relationship between the search term and
24 those sponsored listings that gives rise to the likelihood of
25 confusion.
282
1 One cannot in our view view -- look at the question of
2 the use of "GEICO" as a keyword by itself as presenting the real
3 question to be decided based on the evidence presented in this
4 case.
5 And as we have attempted to make clear, it is the two
6 categories of sponsored listings that at least thus far are
7 generated by GEICO as a keyword that gives rise to the likelihood
8 of confusion. One of those categories is the bulk of the
9 sponsored listings that the searches that we have put into
10 evidence demonstrates, and that is those that include "GEICO" in
11 the text, the title or heading of the sponsored listing.
12 As to that category of sponsored listings, as we said in
13 our opening statement, it seems to me that there is a complete
14 sufficiency of proof, particularly when there is no proof to the
15 contrary that will be offered by Google to conclude that there is
16 a likelihood of confusion, and I say that for two reasons.
17 One is simply by looking at the searches themselves that
18 had -- that establish a connection between a search term and a
19 sponsored listing. Secondly, Dr. Ford's survey, if it proves
20 nothing else, certainly establishes based on the sponsored
21 listings that he tested, both that included the "GEICO" as a
22 search term and the one that did not, that there is an
23 overwhelming likelihood of confusion that is associated with
24 sponsored listings that include the trademark in the text.
25 That is something that's established by his survey.
283
1 It's established by the other evidence submitted, the searches
2 themselves, and particularly given the evidence that came in with
3 respect to GEICO's focal point of its advertising of its business
4 on GEICO as a, a trademark that is used to prompt potential
5 customers to search for a rate quote from GEICO, the fact that a
6 GEICO rate quote cannot be found, the evidence shows, from any
7 site other than GEICO's, and the fact that a consumer would be
8 misled into believing that a third-party source that is offering a
9 comparison of rate quotes that either refers to GEICO specifically
10 or does not refer to GEICO specifically suggests to the consumer
11 that they can find a rate quote from GEICO at those sites, and
12 it's clear, I think, from the evidence that they cannot and that
13 those ads are inherently misleading.
14 Now, Mr. Page refers to the fact that Google already --
15 in its policy already provides that sponsored listings that have
16 the trademark in the text should not be displayed, and as we
17 pointed out, that has always been the case under Google's policy.
18 Its policy did not change with respect to that.
19 And I think the evidence is clear that that has
20 happened. It has continued to happen, and despite the policy, the
21 reality is that those sponsored listings have been displayed
22 despite the objection of GEICO, and in evidence are those
23 objections, in evidence are the numerous instances in which
24 sponsored listings containing the search term "GEICO" continued to
25 appear.
284
1 So far from slipping through, it seems to me that the
2 bulk of the search terms that are in evidence establish that those
3 are the type of categories of listings, that is, with "GEICO" in
4 the title, that have, in fact, occurred in the marketplace and, we
5 believe, clearly give rise to a likelihood of confusion.
6 With respect to the question of whether Dr. Ford's
7 survey establishes a likelihood of confusion as to those sponsored
8 listings that do not have "GEICO" in the title, we submit that
9 Dr. Ford's survey is sufficient to prove that proposition. He
10 did, in fact, test both sponsored listings that include and do not
11 include "GEICO" in the text. The result -- the respondents in the
12 survey were asked independently to react to each of those five
13 sponsored listings.
14 The evidence shows that the responses with respect to a
15 source affiliation were not appreciably different among those
16 respondents. There was a slight dimunition in the level of
17 confusion associated with the fifth sponsored link, but it's still
18 in excess of 50 percent.
19 So we believe that the evidence clearly establishes that
20 there is a likelihood of confusion based both upon the use of
21 "GEICO" in the text, and in this case, given the fact that "GEICO"
22 is a mark associated with rate quotes, that it is also sufficient
23 to establish a likelihood of confusion with respect to those
24 sponsored listings that do not contain "GEICO" in the text.
25 Thank you, Your Honor.
285
1 THE COURT: All right. Well, since this is a bench
2 trial, I could be wrong, but I'm going to assume that this motion
3 is raised under rule 52(c). I think that's the right rule.
4 MR. PAGE: That is correct, Your Honor.
5 THE COURT: And what the Court is being called upon to
6 do at this point in the trial is to evaluate the strength of the
7 plaintiff's evidence to see whether or not the case should go
8 forward on any or all of the issues in this case, and having
9 carefully considered the plaintiff's case, the Court is going to
10 grant in part and deny in part the defendant's motion.
11 As we all know, there are five essential elements to a
12 Lanham Act claim. First, the plaintiff must establish that it
13 possesses a mark that is protectable, and of course, that's not an
14 issue in this case. "GEICO" clearly is.
15 The remaining elements require basically a focus on what
16 the defendant does. Question No. -- the second element is that
17 the defendant uses the mark; three, that the defendant's use of
18 the mark in commerce; four, that the defendant's use of the mark
19 is in connection with the sale, offering for sale, distribution,
20 or advertising of goods and services; and fifth and what has in my
21 mind been the key issue in why I could not grant summary judgment
22 initially in this case or grant the motion to dismiss that was
23 previously filed was that the use of the mark is done in a manner
24 that's likely to cause confusion.
25 Having heard the plaintiff's case, the Court is
286
1 satisfied that the plaintiff has not established that the mere use
2 of its trademark by Google as a search word or keyword or even
3 using it in their AdWord program standing alone violates the
4 Lanham Act because that activity in and of itself, there's no
5 evidence that that activity standing alone causes confusion.
6 The Court also finds that there was insufficient
7 evidence presented in the plaintiff's case to let this case go
8 forward on the question of whether Google violated the Lanham Act
9 after it barred -- or after it began to bar the use of the GEICO
10 mark from either the titles or the text of the sponsored ads that
11 appear as a result of use of the AdWord program.
12 And the reason I find insufficient evidence of that, as
13 we sort of -- I sort of indicated during the testimony of
14 Dr. Ford's, is that, frankly, Dr. Ford either wasn't asked or
15 chose not to actually query that particular issue because the
16 survey focused on either ads that had "GEICO" in the title -- I
17 mean, the page that people were looking at had five sponsored ads,
18 the first four of which had "GEICO" either in the title and/or in
19 the text. The fifth one did not, but it was on a page that had
20 the four previous sponsored links.
21 The test -- the control in this case used Nike, but what
22 I don't feel was presented to the Court that needed to be
23 presented to the Court would have been ads of insurance that did
24 not have "GEICO" in it, and I was not satisfied therefore that the
25 plaintiff's evidence was sufficient.
287
1 However, as to the narrow issue of whether the plaintiff
2 has presented sufficient evidence at this point to let the case
3 continue on the issue of whether Google is liable for violating
4 the Lanham Act for those sponsored ads using GEICO's name in
5 either the title or the text that appear next to a -- and a GEICO
6 organic listing as a result of the AdWord program, at this point,
7 there has been enough evidence of confusion as a result of
8 Dr. Ford's report to deny the motion and to allow the case to
9 continue so the Court can consider the defendant's evidence as to
10 whether or not this situation creates a likelihood of confusion in
11 the marketplace and, if so, what damages might be appropriate.
12 Now, counsel, is my ruling clear, and any questions
13 about what I have just done?
14 MR. OSSOLA: Your Honor, I have just one.
15 THE COURT: Yes.
16 MR. OSSOLA: You made reference to the -- what you
17 believed to be the insufficiency of evidence after Google began to
18 bar the use of the trademark in the sponsored link. I think
19 that's what you said. I just want to make clear that Google has
20 always barred --
21 THE COURT: And I should clarify that. I realize that,
22 but in this particular case, they -- I meant to confine that to
23 the facts of this case; that is, I didn't mean to suggest that
24 Google did not have that policy in place, but the reality was that
25 we do have evidence that there were -- "GEICO" was being used in
288
1 titles and in text until Google was advised to stop that, and then
2 following its own policy, it implemented that policy.
3 MR. OSSOLA: I understand.
4 THE COURT: All right.
5 MR. OSSOLA: But the insufficiency evidence on that
6 point is that with respect to those sponsored listings that do not
7 have "GEICO" in the text or heading of the ad, as to those -- that
8 category of sponsored listings, the Court is finding that there is
9 not sufficient evidence of likelihood of confusion.
10 THE COURT: In this case.
11 MR. OSSOLA: In this case.
12 THE COURT: Correct.
13 MR. OSSOLA: Thank you.
14 THE COURT: All right?
15 Mr. Page, was there anything you needed clarified?
16 MR. PAGE: No, I think that covers it, Your Honor.
17 THE COURT: All right. Now, I think as we all know,
18 there's serious legal issues that are involved in this decision
19 that I've just rendered, in particular, the status of how
20 trademarks as used in keywords, etc., and I think I would like
21 time to write a more detailed opinion on these legal issues, and
22 so depending upon how you-all want to proceed, as you know, I've
23 also been encouraging you to see if you can resolve the case, and
24 I've given you now some very clear parameters as to how the case
25 would continue.
289
1 Unless there's any objection, what I propose is that we
2 terminate the trial at this point, I don't mean end it, but stop
3 right now what we're doing to give the Court a brief amount of
4 time, which given the holiday season might be two or three weeks,
5 although we're going to try to do it sooner than that, to get a
6 written opinion out on this ruling, consistent with this ruling,
7 and to allow you-all the opportunity to see whether or not there
8 can be a resolution of what is left in the case.
9 Does that meet with all of your approval?
10 MR. PAGE: Yes, Your Honor, that would be fine.
11 MR. OSSOLA: Yes, Your Honor.
12 THE COURT: All right. Now, I'm not going to set
13 another date at this point for when we would start up Round 2 if
14 that becomes necessary. We'll get the opinion out as quickly as
15 possible, and then you can get back to me as to how you want to
16 proceed.
17 In the meantime, since this courtroom is not going to be
18 used again until January 3, and I'm hoping we have it resolved by
19 then, you can if you want leave your exhibits up here. If the
20 case ultimately has been resolved, then there's no need for any of
21 the exhibits to stay at the courthouse, and we'll ask you to come
22 and get them.
23 If the case goes on to the, what I would call the second
24 phase, then obviously, we'll need to get the appellate record
25 cleaned up in terms of just what exhibits are in or out. All
290
1 right?
2 MR. OSSOLA: Yes.
3 THE COURT: Anything further we need to address at this
4 point?
5 MR. OSSOLA: Your Honor, just the second phase, if we
6 get there, would be -- would include damages as well; is that
7 correct?
8 THE COURT: Yes. It would be the defendant's
9 opportunity to rebut liability, and obviously, if I were to find
10 that as a matter of law there was no issue left on liability and
11 find in the defendant's favor, then we wouldn't have to have a
12 damages phase, but if I find liability, then there would be a
13 damages phase. All right?
14 MR. OSSOLA: What's been established to date then is
15 that the category of sponsored listings that have "GEICO" in the
16 title and text are -- have been found to be confusing and
17 misleading under the Lanham Act?
18 THE COURT: No. I have found that you've presented
19 enough evidence at this point to avoid a motion for judgment as a
20 matter of law. In other words, there certainly is a prima facie
21 case that you've established that they are confusing based on
22 Dr. Ford's survey.
23 Now, I haven't heard the rebuttal evidence, so I'm not
24 making a specific finding on that. I'm at this point weighing the
25 quality of the evidence, the amount of evidence. You've presented
291
1 enough to let the case go forward.
2 MR. OSSOLA: Thank you, Your Honor.
3 THE COURT: All right?
4 MR. PAGE: Thank you, Your Honor.
5 THE COURT: All right? And, counsel, that one piece of
6 paper that was on each of your chairs was a joke. It was not
7 serious. But we happened to see it and thought it was
8 appropriate.
9 All right, I wish you-all a happy holiday, and we'll get
10 back to you as quickly as possible.
11 We'll recess court for the day.
12 MR. OSSOLA: Thank you, Your Honor.
13 MR. PAGE: Thank you, Your Honor.
14 (Recess from 10:31 a.m., until 11:09 a.m.)
15 THE COURT: All right, for the record, Mr. Ossola has
16 brought to the Court's attention the following matter: My
17 understanding, Mr. Ossola, is that your view of the defendant's
18 evidence is that Google will not be able to offer any evidence to
19 counter GEICO's evidence that those sponsored sites that contain
20 "GEICO" either in their title or in their text that have been at
21 issue in this case are confusing. Is that correct?
22 MR. OSSOLA: That's correct, Your Honor.
23 THE COURT: And it's my understanding that, what, the
24 expert -- is it Dr. Jacoby, Mr. Page? Whose report would have --
25 MR. PAGE: That's correct. It would be Dr. Jacoby's
292
1 report. Mr. Ossola is correct. Our survey addressed Google's
2 current policy, so the sponsored links that we tested do not
3 contain the word "GEICO." We were testing the keyword issue.
4 THE COURT: All right.
5 MR. PAGE: So Dr. Jacoby's survey says nothing either
6 way as to whether a link with GEICO is confusing.
7 THE COURT: All right. Then the Court should have
8 clarified its earlier -- well, it couldn't clarify -- should have
9 said in its earlier ruling and says now in response to your motion
10 that the evidence before this Court does establish that those
11 sponsored sites that contain "GEICO" either in the title or in the
12 text are likely to confuse for purposes of the Lanham Act
13 requirements, and therefore, to the extent that the defendant's
14 motion was based on that issue, that will also be denied.
15 And that reduces the issues that would have to go
16 forward were there -- if there is a second phase to this trial.
17 The sole issues that would remain -- and correct me if I'm
18 wrong -- are, No. 1, whether or not Google is liable for any
19 Lanham Act violation based upon those sponsored sites, and 2, if
20 Google were liable, then what damages would be appropriate.
21 Now, is that a correct phrasing from your standpoints as
22 to what would be left in this case?
23 MR. PAGE: That is correct, Your Honor. The issue that
24 will remain is, is since the Court has found a likelihood of
25 confusion for ads that have "GEICO" in them, the question is
293
1 whether Google is contributorily liable or whether only the
2 advertiser would be liable.
3 THE COURT: Well, actually, because the advertisers are
4 not in this case, the only question before us would be whether or
5 not Google is liable.
6 MR. PAGE: That's correct, Your Honor.
7 THE COURT: Right. Mr. Ossola?
8 MR. OSSOLA: I believe that's right. Under contributory
9 liability or any other theory that may be asserted by GEICO, that
10 remains to be seen.
11 THE COURT: Correct.
12 MR. OSSOLA: Phase 2.
13 THE COURT: So there has been -- just so we're clear for
14 the record, there has been no finding of liability -- there's been
15 no finding that Google is liable at this point. What we have
16 found, however, is that this particular group of sponsored sites
17 does violate the Lanham Act.
18 MR. OSSOLA: Yes, Your Honor.
19 THE COURT: All right?
20 MR. PAGE: That's correct.
21 THE COURT: Do you think we need to clarify anything
22 further for the record?
23 MR. OSSOLA: No, Your Honor, thank you.
24 THE COURT: All right. Thank you very much for bringing
25 that to our attention. We'll recess court once again.
294
1 (Recess at 11:12 a.m.)
2
3 CERTIFICATE OF THE REPORTER
4 I certify that the foregoing is a correct transcript of the
5 record of proceedings in the above-entitled matter.
6
7
8
Anneliese J. Thomson
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
295
1 I N D E X
2 DIRECT CROSS REDIRECT RECROSS
3 WITNESS ON BEHALF OF
THE PLAINTIFF:
4
Gary T. Ford, Ph.D. 271
5 (Resumed)
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25