J.T. Colby & Company, Inc. et al v. Apple, Inc.

Filing 121

MEMORANDUM OF LAW in Support re: 120 MOTION in Limine to Exclude any Testimony, Argument or Evidence Regarding the Expert Reports and Opinions of Susan Schwartz McDonald. REDACTED. Document filed by Apple Inc.. (Cendali, Dale)

Dale Cendali
Claudia Ray
Bonnie L. Jarrett
KIRKLAND & ELLIS LLP
601 Lexington Avenue
New York, New York 10022

Perry J. Viscounty
LATHAM & WATKINS LLP
140 Scott Drive
Menlo Park, CA 94025
Jennifer L. Barry
LATHAM & WATKINS LLP
600 West Broadway, Suite 1800
San Diego, CA 92101-3375

Attorneys for Defendant
APPLE INC.
UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
J.T. COLBY & COMPANY, INC. d/b/a BRICK
TOWER PRESS, J. BOYLSTON & COMPANY,
PUBLISHERS LLC and IPICTUREBOOKS LLC,
Plaintiffs,

Case No. 11-CIV-4060 (DLC)
ECF Case

- against APPLE INC.,

REDACTED
Defendant.

MEMORANDUM OF LAW IN SUPPORT OF DEFENDANT’S MOTION TO
EXCLUDE ANY TESTIMONY, ARGUMENT OR EVIDENCE REGARDING
THE EXPERT REPORTS AND OPINIONS OF SUSAN SCHWARTZ McDONALD

TABLE OF CONTENTS
Page
PRELIMINARY STATEMENT ..................................................................................................1
ARGUMENT ..................................................................................................................................4
I.

LEGAL STANDARD. .......................................................................................................4

II.

DR. MCDONALD’S EXPERT REPORTS ARE BIASED. ...........................................5

III.

THE SURVEYS FAILED TO REPLICATE MARKET CONDITIONS. ...................6
A.

The “Conceptual” Stimulus Did Not Replicate Market Conditions. .......................7
1.

B.

The “Conceptual” Stimulus Violated A Fundamental Requirement
Of Consumer Surveys. .................................................................................7

The “Conceptual” Stimulus Biased The Results By Depriving
Respondents Of Marketplace Clues. ........................................................................9
1.
2.

C.

The McDonald Surveys Did Not Depict The “ibooks” Logo. ...................10
The McDonald Surveys Did Not Include Any Information About
The Copyright Or Title Page, Or The Appearance Of Plaintiffs’
Imprint........................................................................................................11

The McDonald Surveys Placed Undue Emphasis on the iBooks Imprint .............13

IV.

THE MCDONALD SURVEYS FAILED TO ASK THE GENERALLY
ACCEPTED QUESTION FOR TESTING CONFUSION. .........................................14

V.

THE MCDONALD SURVEYS FAILED TO USE A PROPER CONTROL ............16

VI.

DR. MCDONALD’S SURVEYS SHOULD BE EXCLUDED AS WELL
BECAUSE HER METHODOLOGY SUFFERED FROM ADDITIONAL
FATAL FLAWS. ..............................................................................................................18
A.

The McDonald Surveys Primed Respondents To Think Of Apple. ......................18

B.

The McDonald Surveys Did Not Test The Proper Universe. ................................19

C.

The McDonald Surveys Did Not Properly Screen Respondents. ..........................20

D.

The McDonald Surveys Failed To Take Steps To Prevent Guessing. ...................21

E.

Dr. McDonald Improperly Calculated The Rate Of Confusion. ............................22

F.
G.
VII.

The McDonald Surveys Were Not Validated. .......................................................23
Dr. McDonald Improperly Speculates About The Future. ....................................24

DR. MCDONALD IMPROPERLY OPINES ON PLAINTIFFS’ AND
APPLE’S BRAND DEVELOPMENT ...........................................................................24

TABLE OF AUTHORITIES
Page
Cases
Am. Footwear Corp. v. General Footwear Co. Ltd.,
609 F.2d 655 (2d Cir. 1979)........................................................................................ 10, 11, 19
AstraZeneca v. Tap Pharmaceutical,
444 F. Supp. 2d 278 (D. Del. 2006) .......................................................................................... 6
Conopco, Inc. v. Cosmair, Inc.,
49 F. Supp. 2d 242 (S.D.N.Y. 1999)....................................................................................... 20
Cumberland Packing Corp. v. Monsanto Co.,
32 F. Supp. 2d 561 (E.D.N.Y. 1999) .................................................................... 16, 17, 21, 23
Daubert v. Merrell Dow Pharmaceuticals, Inc.,
509 U.S. 579 (1993) .......................................................................................................... 1, 4, 5
General Motors Corp. v. Lanard Toys, Inc.,
468 F.3d 405 (6th Cir. 2006) ............................................................................................ 16, 22
Inc. Publ’g Corp. v. Manhattan Magazine, Inc.,
616 F. Supp. 370 (S.D.N.Y. 1985), aff’d sub nom. Inc. Publ’g Corp. v.
Manhattan Magazine, 788 F.2d 3 (2d Cir. 1986) ............................................................... 9, 14
Juicy Couture, Inc. v. L'Oreal USA, Inc.,
04 CIV.7203(DLC), 2006 WL 1012939 (S.D.N.Y. Apr. 19, 2006) ............................. 9, 13, 18
Kargo Global, Inc. v. Advance Magazine Publishers, Inc.,
06 CIV 550 JFK, 2007 WL 2258688 (S.D.N.Y. Aug. 6, 2007) ................................. 5, 7, 9, 13
Lippe v. Bairnco Corp.,
288 B.R. 678, 686 (S.D.N.Y. 2003), aff'd, 99 F. App'x 274 (2d Cir.
2004) ................................................................................................................................. 24, 25
Malletier v. Dooney & Bourke, Inc.,
525 F. Supp. 2d 558 (S.D.N.Y. 2007).............................................................................. passim
Mastercard Int'l Inc. v. First Nat. Bank of Omaha, Inc.,
02 CIV. 3691 (DLC), 2004 WL 326708 (S.D.N.Y. Feb. 23, 2004) ............................. 5, 21, 23
Schering Corp. v. Pfizer Inc.,
189 F.3d 218 (2d Cir. 1999)...................................................................................................... 5
Starter Corp. v. Converse, Inc.,
170 F.3d 286 (2d Cir. 1999)...................................................................................................... 5

TABLE OF AUTHORITIES (cont’d)
Page
THOIP v. Walt Disney Co.,
690 F. Supp. 2d 218 (S.D.N.Y. 2010).............................................................................. passim
THOIP v. Walt Disney Co.,
788 F. Supp. 2d 168 (S.D.N.Y. 2011)............................................................................... 22, 23
Union Carbide Corp. v. Ever-Ready, Inc.,
531 F.2d 366 (7th Cir. 1976) .................................................................................................. 15
Universal City Studios, Inc. v. Nintendo Co., Ltd.,
746 F.2d 112 (2d Cir. 1984).................................................................................................... 19
Vista Food Exch., Inc. v. Vistar Corp.,
03-CV-5203DRHWDW, 2005 WL 2371958 (E.D.N.Y. Sept 27, 2005) ........................... 7, 14
WE Media, Inc. v. Gen. Elec. Co.,
218 F. Supp. 2d 463 (S.D.N.Y. 2002), aff’d sub nom. WE Media, Inc.
v. Cablevision Sys. Corp., 94 F. App’x 29 (2d Cir. 2004) ............................................... passim
Weiner v. Snapple Beverage Corp.,
07 CIV. 8742 DLC, 2010 WL 3119452 (S.D.N.Y. Aug. 5, 2010) ......................................... 25
Rules
Fed. R. Evid. 403 .................................................................................................................... 1, 4, 5
Fed. R. Evid. 702 .................................................................................................................... 1, 4, 5
Other Authorities
Expert Report of Susan Schwartz McDonald, Ph.D., PBM Products, LLC
v. Societe Des Produits Nestle, S.A., Case No. 3:09-CV-0030
(W.D.Vir. 2009) ........................................................................................................................ 9
Federal Judicial Center, Reference Manual of Scientific Evidence, 3rd ed.
(2011) ........................................................................................................................ 5, 9, 14, 23
G. Kip Edwards, The Daubert Revolution and Lanham Act Surveys,
Trademark and Deceptive Advertising Surveys ....................................................................... 9
Jacob Jacoby, Experimental Design and the Selection of Controls in
Trademark and Deceptive Advertising Surveys, Trademark Rep., Vol.
92 (2002) ................................................................................................................................. 16
Jerre B. Swann, Likelihood of Confusion, Trademark and Deceptive
Advertising Surveys 76 (2012) ........................................................................................... 7, 15

TABLE OF AUTHORITIES (cont’d)
Page
McCarthy on Trademarks and Unfair Competition (4th ed. 2012) .............................. 6, 14, 15, 23
Shari Seidman Diamond, Control Foundations: Rationale and
Approaches, Trademark and Deceptive Advertising Surveys (2012) .................................... 17
Soaring iPhone 5 sales in US knock Android into second place, Kantar
World Panel (Nov. 11, 2012),
http://www.kantarworldpanel.com/global/News/Soaring-iPhone-5sales-in-US-knock-Android-into-second-place) ..................................................................... 18
William G. Barber, The Universe, Trademark and Deceptive Advertising
Surveys (2012) .................................................................................................................. 20, 21

Defendant Apple Inc. (“Apple”) respectfully submits this memorandum of law in support
of its motion to exclude all testimony, argument or evidence relating to the opinions and reports
of Dr. Susan McDonald, which were submitted on September 17, 2012 and December 6, 2012,1
and may be offered by Plaintiffs J.T. Colby & Co., Inc. d/b/a Brick Tower Press, J. Boylston &
Co., Publishers LLC, and Ipicturebooks LLC (collectively, “Plaintiffs”).
PRELIMINARY STATEMENT
If ever the Court should exclude a survey, it should be the two designed and conducted
by Dr. McDonald in this case. Her on-line surveys are so profoundly flawed and biased that they
come nowhere near meeting the requirements of reliability or relevancy under Daubert v.
Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). Thus, the surveys and any testimony,
argument or evidence relating to them should be excluded under Fed. R. Evid. 702 and 403.
In designing her first online likelihood of confusion survey, Dr. McDonald committed the
most basic of mistakes: she tested the wrong mark. A reverse confusion survey should test
consumer responses to a stimulus showing the alleged senior user’s mark. But her first survey
not only failed to use any stimulus, rather it asked respondents to “envision” a page in an
electronic book with the word “iBooks” on it—emulating the formulation Apple uses for its
famous family of i-formative marks rather than the one Plaintiffs use for their “ibooks” imprint.
As stated in Plaintiffs’ Amended Complaint, their mark is “ibooks,” in all lowercase letters. In
fact, Plaintiffs did not begin depicting their “ibooks” imprint with a lowercase “i” and capital
“B,” like Apple’s i-formative marks, until after Apple announced its iBooks application in
January 2010—a change apparently made by Plaintiffs either to help their case in this litigation
1

Dr. McDonald’s first report, A Survey to Measure Potential Source Confusion Associate with iBooks, submitted on
September 17, 2012, is annexed to the Declaration of Mary Mazzello dated December 21, 2012 (“Mazzello Dec.”)
as Exhibit 9. Dr. McDonald’s sur-rebuttal report, Sur-rebuttal and Supplemental Report of Susan Schwartz
McDonald Ph.D., submitted on December 6, 2012 (the “Sur-Rebuttal Report”) is annexed to the Mazzello Dec. as
Exhibit 10. Defendants will cite directly to these expert reports rather than the declaration.

or to improperly take advantage of Apple’s goodwill (Mazzello Dec. ¶ __.) Because Dr.
McDonald’s first survey tested the wrong mark, it does not assess probative reverse confusion.
After viewing the report of Apple’s rebuttal survey expert, Dr. Stephen M. Nowlis,
Distinguished Professor of Marketing in the Olin Business School at Washington University in
St. Louis (the “Nowlis Report”), which highlighted this error, Dr. McDonald conducted yet
another survey, changing only the reference in the on-line questionnaire to “ibooks,” and keeping
all other aspects of the survey the same. The Sur-Rebuttal Survey in any case was in violation
of this Court’s scheduling order, which did not provide for service of sur-rebuttals.
Although the failure to test the right alleged mark alone is dispositive, Dr. Nowlis’ report
also details an unusual number of additional deep-rooted flaws in Dr. McDonald’s methodology
that permeate both surveys. Significantly, even Dr. Jacob Jacoby, Plaintiffs’ own survey rebuttal
expert, repeatedly agreed with Dr. Nowlis’ opinion that the types of methodological choices
made by Dr. McDonald were improper. In particular, Dr. Jacoby agreed with Dr. Nowlis’
opinion on each of the following three fatal flaws, on which we believe the Court should focus.
First, Dr. McDonald made no effort to replicate real-world market conditions — a key
requirement for any consumer survey. Indeed, in both surveys she instructed respondents to
“envision” a page in a digital book rather than showing them a visual stimulus that replicated the
manner in which consumers would see Plaintiffs’ alleged mark in the real world. Dr. McDonald
thus tested confusion in the abstract, devoid of context and the source identifying information
that she admitted would typically appear within Plaintiffs’ books.
Second, instead of asking a variation on the time-tested standard question, “Who makes
or puts out this product?” Dr. McDonald asked respondents, “Who makes this product
available?” This formulation leads to worthless results because “available” is a broad concept.

2

Respondents could correctly answer “Apple” in response to Dr. McDonald’s question as Apple
does in fact make third party content such as electronic books available through its iBookstore,
its iPad and iPhone devices, and its iTunes store, without this evidencing any confusion.
Third, although Dr. McDonald admitted that consumers associate “i”-formative marks
with Apple, she did nothing to control for people who answered “Apple” simply because of the
presence of “i” in the word “ibooks.” Plaintiffs do not claim any exclusive rights in the letter “i,”
standing alone, only in “i” plus “books.” Accordingly, the control should have had an “i” prefix,
as did those in the surveys conducted by Apple’s experts, Dr. E. Deborah Jay and Dr. Nowlis.
Instead, Dr. McDonald used the generic term, “eBooks” as the control.
Although each of these flaws, standing alone, renders Dr. McDonald’s contrived surveys
valueless, there are numerous other serious errors that also mandate exclusion, including:
•

Dr. McDonald skewed the results by gratuitously mentioning “smartphones” in a screener
question immediately before asking the main survey questions, improperly conditioning
respondents to think of Apple, maker of the widely popular iPhone smartphones.

•

Dr. McDonald surveyed the wrong universe, surveying only those who had downloaded
(not purchased) a digital book in the past. Thus, she did not screen for future behavior —
i.e., prospective purchasers — nor did her universe include purchasers of digital books.

•

Dr. McDonald’s universe was materially under-inclusive — respondents had to have
downloaded a digital book in the last six months, despite the fact that
of
Plaintiffs’ sales are for print, not digital, books.

•

Dr. McDonald failed to screen out “atypical” respondents, such as (1) people who work
in publishing, advertising, or market research and (2) people who are “professional”
survey takers, all of whom should have been excluded because they have special
knowledge that could skew survey results.

•

Dr. McDonald strayed from standard survey methodology by failing to tell respondents
not to guess—and the verbatim responses plainly showed substantial guessing occurred.

•

Dr. McDonald calculated the survey results to maximize a finding of confusion by
classifying as confused respondents who named Apple solely because of the letter “i” or
gave reasons unrelated to the alleged mark “ibooks.” When properly calculated, despite

3

the survey’s numerous flaws, the confusion levels would be 5.2% for her initial survey
and 3.8% for her Sur-Rebuttal survey—far below the threshold for actionable confusion.
•

Dr. McDonald failed to validate her surveys, as is standard procedure.

•

As further proof of her unusual, nonobjective approach, Dr. McDonald said she could
“easily offer the opinion that the level of source confusion would be even larger in a
study conducted several years from now”—a nonsensical and unscientific opinion that
cannot possibly be supported as it obviously is impossible to test the future.
Many of these flaws, in and of themselves, would be an appropriate basis for excluding

Dr. McDonald’s testimony. Taken together, they demonstrate that Dr. McDonald’s surveys,
reports and opinions are so fatally flawed, and deviate from accepted methodology in such
significant and numerous respects, that her surveys, report and opinions, and any testimony,
evidence or argument relating to them, should be excluded in their entirety.
ARGUMENT
I.

LEGAL STANDARD.
Expert testimony is admissible only if “(1) the expert’s scientific, technical, or other

specialized knowledge will help the trier of fact to understand the evidence or to determine a fact
in issue; (2) the testimony is based upon sufficient facts or data, (3) the testimony is the product
of reliable principles and methods, and (4) the expert has reliably applied the principles and
methods to the facts of the case.” Fed. R. Evid. 702. Federal Rule of Evidence 403, which
serves as yet another check on the admissibility of expert evidence, states that “[t]he court may
exclude relevant evidence if its probative value is substantially outweighed by a danger of . . .
unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or
needlessly presenting cumulative evidence.” Fed. R. Evid. 403; see also Daubert, 509 U.S. at
589 (to be admissible, expert testimony must be “not only relevant, but reliable”).
Courts serve as the “gatekeeper” of expert evidence. See Malletier v. Dooney & Bourke,
Inc., 525 F. Supp. 2d 558, 561-62 (S.D.N.Y. 2007)). Because expert testimony has the potential
4

to be particularly misleading, courts must exercise greater control over experts than over lay
witnesses. Daubert, 509 U.S. at 595. Stricter scrutiny is also warranted where, as here, a jury
trial rather than a bench trial is contemplated. See Kargo Global, Inc. v. Advance Magazine
Publishers, Inc., 06 CIV 550 JFK, 2007 WL 2258688, at *6, 11 (S.D.N.Y. Aug. 6, 2007).
As discussed below, the McDonald surveys are singularly unreliable and prejudicial, and
thus should be excluded. See Starter Corp. v. Converse, Inc., 170 F.3d 286, 296-97 (2d Cir.
1999) (affirming district court’s exclusion of survey where prejudicial effect outweighed
probative value); THOIP v. Walt Disney Co., 690 F. Supp. 2d 218, 241 (S.D.N.Y. 2010)
(excluding survey because it was not reliable); Malletier, 525 F. Supp. at 568-75 (excluding
surveys as unreliable and irrelevant); Kargo, 2007 WL 2258688, at *12 (excluding survey where
its prejudicial effect and the likelihood that it would confuse the jury substantially outweighed its
probative value); Mastercard Int'l Inc. v. First Nat. Bank of Omaha, Inc., 02 CIV. 3691 (DLC),
2004 WL 326708, at *10 (S.D.N.Y. Feb. 23, 2004) (excluding survey under Rules 702 and 403).
II.

DR. MCDONALD’S EXPERT REPORTS ARE BIASED.
As an initial matter, the McDonald surveys fail to meet the most basic requirement of any

scientific study, that it be “based on principles of hypothesis generation, scrupulous study design,
meticulous data collection, and objective interpretation of experimental results.” Federal Judicial
Center, Reference Manual of Scientific Evidence, 3rd ed., at xiii (2011) (“Reference Manual”);
see also Mastercard, 2004 WL 326708, at * 8 (one “factor[] courts consider in assessing the
reliability of a survey [is] whether . . . the objectivity of the entire process was ensured”; failure
to satisfy this criteria may result in the exclusion of the survey) (quoting Schering Corp. v. Pfizer
Inc., 189 F.3d 218, 225 (2d Cir. 1999)); Mazzello Dec., Ex. 12 (Jacoby Dep. 23:9-15.). On the
very first page of her first report, Dr. McDonald admits that the “survey confirms [her] opinion
that . . . Apple’s mark creates significant confusion at present and potentially greater confusion in
5

the future . . . .” (McDonald Rep. at 1.) In other words, she admittedly designed her first survey
to substantiate her pre-existing point of view. In fact, she admitted that as early as 2010, after a
single conversation with Plaintiffs’ prior counsel, “[she] thought that, in fact, there could be a
very real likelihood of confusion . . . .” (Mazzello Dec., Ex. 13 (McDonald Dep. 22:23-23:25.).)
Dr. McDonald’s anti-Apple bias infects her expert opinion and her surveys. She refers to
Apple as an “i-imperialis[t]” who “commandeered the ‘i’ prefix for its book distribution
business” and claims that Apple “coopted” the word “ibooks.” (McDonald Rep. at 2, 5, 7.) She
also opines on Apple’s intent—as if she were litigation counsel—claiming that it “aspires to own
the prefix i,” is “seeking to dominate virtually every arena and conduit of digital consumption,”
and will pursue its branding strategy “even at the risk of trespassing on other brands that might
stand in its way.” (McDonald Rep. at 5.) But Apple’s purported intent is not even close to being
within Dr. McDonald’s realm of expertise. She, of all people, should understand this, having had
her supplemental expert report in AstraZeneca v. Tap Pharmaceutical excluded because she
improperly opined on the opposing party’s intent. See 444 F. Supp. 2d 278, 293-94 (D. Del.
2006). As discussed below, this palpable anti-Apple bias infected her entire survey design and
likely led to the multitude of flaws that, individually and collectively, render the McDonald
surveys unreliable, prejudicial and inadmissible. See Malletier, 525 F. Supp. 2d at 569-70; 603
(excluding survey that was neither conducted nor reported in an objective manner).
III.

THE SURVEYS FAILED TO REPLICATE MARKET CONDITIONS.
A survey must “compare the impressions the marks have on potential customers under

marketplace conditions.” WE Media, Inc. v. Gen. Elec. Co., 218 F. Supp. 2d 463, 474 (S.D.N.Y.
2002), aff’d sub nom. WE Media, Inc. v. Cablevision Sys. Corp., 94 F. App’x 29 (2d Cir. 2004);
see also 6 McCarthy on Trademarks and Unfair Competition § 32:163 (4th ed. 2012) ( survey
must “mirror the situation in which the ordinary person would encounter the trademark”). The
6

failure to approximate actual marketplace conditions is grounds for inadmissibility. See THOIP,
690 F. Supp. 2d at 241 (excluding survey that, inter alia, did not replicate market conditions);
Malletier, 525 F. Supp. 2d at 595 (improper stimulus was “enough on its own to justify
exclusion”); Kargo, 2007 WL 2258688, at *7 (excluding survey that, inter alia, did not replicate
market conditions); Vista Food Exch., Inc. v. Vistar Corp., 03-CV-5203DRHWDW, 2005 WL
2371958, at *5 (E.D.N.Y. Sept 27, 2005) (same). Having utterly failed to replicate marketplace
conditions, the McDonald surveys should be excluded for this reason alone.
A.

The “Conceptual” Stimulus Did Not Replicate Market Conditions.
1.

The “Conceptual” Stimulus Violated A Fundamental Requirement Of
Consumer Surveys.

As even Dr. McDonald admits, the “‘standard’ approach” to a likelihood of confusion
survey “involves selection of a stimulus that can plausibly represent market exposure . . . .”
(McDonald Rep. at 4; see id. at 1.) It is fundamental that a survey must use a stimulus “‘that
approximate[s] what a potential customer would encounter in making’ purchasing decisions.”
Kargo, 2007 WL 2258688, at *10 (quoting WE, 218 F. Supp. 2d at 474) (“Typically, trademark
infringement surveys use stimuli, such as pictures, advertisements or clothing, that directly
expose potential consumers to the products or the marks in question.” (citation omitted)); see
also Jerre B. Swann, Likelihood of Confusion, in Trademark and Deceptive Advertising Surveys
76 (2012) (“confusion cannot be assessed in the abstract”); Mazzello Dec., Ex. 12 (Jacoby Dep.
195:11-16) (“it would be wrong just to look at the mark in isolation, devoid of context”).
Dr. McDonald did not even attempt to comply with this requirement. (See Mazzello
Dec., Ex. 13 (McDonald Dep. 121:22-122:10.).) She did not show respondents the “ibooks”
imprint on any print books, on any pages from electronic books, or on any images from websites
where books can be purchased, although these stimuli reflect the actual market conditions in
7

which Plaintiffs’ books are purchased and were readily available. (See id. (McDonald Dep.
135:19-25; 149:2-22.).) In fact, she did not use a visual stimulus of any kind. Instead, she
admittedly conducted an “unusual survey” (id. (McDonald Dep. 146:5)) and used what she
described as a “conceptual stimulus” (McDonald Rep. at 2), whereby survey respondents were
presented with the following prompt instructing them to “envision” a page in a digital book:
Q1a. Please envision the following scenario, involving a digital/electronic book.
In the scenario we’d like you to envision, you are looking at the particular “page” of a
digital/electronic book that contains information about the book—such as the date of
publication, the publisher, the Library of Congress number, etc.
If, on that page, you see the word [“iBooks” / “eBooks”]; [“ibooks” / “ebooks”] what
company or companies would you think had made the book available?
(McDonald Rep. at 11, Sur-Rebuttal Rep. at 3.) No other details about the book, the alleged
mark, or the other information on the conceptual page were provided.
Dr. McDonald claimed a conceptual stimulus is proper because one cannot simulate “the
market encounter between consumer end [sic] product . . . in a meaningful way and at the same
time fully do justice to the way that consumers ultimately apprehend a so-called publisher or
source of a book,” (Mazzello Dec., Ex. 13 (McDonald Dep. 136:17-137:4.) This makes no
sense. There is no basis for treating Plaintiffs’ books differently from the myriad goods that
have been tested using actual stimuli—especially given the certainty that a consumer would
never see the “ibooks” imprint on a page without any source-identifying information. (See
Mazzello Dec., Exs. 5-8.) In any case, Dr. McDonald could not indentify any courts that have
blessed the use of a “conceptual stimulus.” (Mazzello Dec., Ex. 13 (McDonald Dep. 157:8-16.)
By employing this bizarre “conceptual” (i.e. imaginary) stimulus, which necessarily
would be “envisioned” differently by each and every respondent, Dr. McDonald utterly failed to
replicate market conditions in even the most basic way — by showing respondents a visual
8

representation of the product bearing the mark in issue. In the few cases where experts have
dared to conduct a survey without using an actual stimulus, courts have routinely held that the
survey is not probative of confusion.2 See e.g. Juicy Couture, Inc. v. L'Oreal USA, Inc., 04
CIV.7203(DLC), 2006 WL 1012939, at *25 (S.D.N.Y. Apr. 19, 2006) (survey not probative of
confusion because, among other things, respondents were shown a card with the trademark on it
rather than a picture or display); WE, 218 F. Supp. 2d at 474 (same); Inc. Publ’g Corp. v.
Manhattan Magazine, Inc., 616 F. Supp. 370, 393-94 (S.D.N.Y. 1985) (holding that telephone
survey had no probative value because, inter alia, respondents were not presented with a visual
stimulus) aff’d sub nom. Inc. Publ’g Corp. v. Manhattan Magazine, 788 F.2d 3 (2d Cir. 1986);
G. Kip Edwards, The Daubert Revolution and Lanham Act Surveys, in Trademark and Deceptive
Advertising Surveys, at 346 (2012) (“Most Lanham Act surveys involve showing survey
respondents something . . .”). Because Dr. McDonald’s surveys had no connection to what
consumers actually observe in the marketplace, they should be excluded as wholly unreliable.
See THOIP, 690 F. Supp. 2d at 241 (excluding survey that did not replicate market conditions
among other things); Kargo, 2007 WL 2258688, at *7 (same).
B.

The “Conceptual” Stimulus Biased The Results By Depriving Respondents
Of Marketplace Clues.

As Dr. McDonald herself has stated, “source inferences reflect a number of contextual
clues integrated by consumers in complex ways.” Expert Report of Susan Schwartz McDonald,
Ph.D., PBM Products, LLC v. Societe Des Produits Nestle, S.A., Case No. 3:09-CV-0030, 14
(W.D.Vir. 2009). Plaintiffs’ other survey expert, Dr. Jacoby, also agrees that “a survey expert
should not deprive respondents of contextual clues that might be helpful to them one way or the
2

Because the “conceptual” stimulus only existed in the mind of each respondent, and because respondents were not
required to describe what they imagined or document their mental images in any way, the McDonald surveys cannot
be standardized or replicated, and thus are inherently unreliable. See Federal Judicial Center, Reference Manual on
Scientific Evidence, at 604 (“The need to replicate research findings permeates most fields of science”).

9

other in assessing confusion.” (Mazzello Dec., Ex. 12 (Jacoby Dep. 197:25-198:5.).)
Nonetheless, here, Dr. McDonald admitted that she “didn’t actually provide [respondents] with
any of the source information that they might find” on one of plaintiffs’ “ibooks” (Mazzello
Dec., Ex. 13 (McDonald Dep. 193:11-16.)). She failed to provide this information even though
she agreed that information such as Plaintiffs’ logo, the name of the publisher, the name of the
author, and the website address serve source-identifying functions (see id. (McDonald Dep.
175:16-182:11.)). Courts have excluded surveys for failing to display source-identifying
information, such as hang tags, that typically appear in the marketplace alongside the product in
issue. See Am. Footwear Corp. v. General Footwear Co. Ltd., 609 F.2d 655, 660 n. 4 (2d Cir.
1979) (affirming rejection of a survey that showed respondents an advertisement that typically
was surrounded by Plaintiffs’ branding in a context devoid of branding); THOIP, 690 F. Supp.
2d at 239 (excluding survey in which expert failed, among other things, to include tags which
would have provided respondents with “source information [that likely] would have diminished
confusion”); WE, 218 F. Supp. 2d at 474 (finding survey that showed respondents words devoid
of context was not probative of confusion and upholding summary judgment for defendant). As
these cases show, comparatively minor deviations from marketplace conditions provide a basis
for excluding a survey — Dr. McDonald’s survey goes far beyond this by showing respondents
nothing at all. This is a low watermark in survey design.
1.

The McDonald Surveys Did Not Depict The “ibooks” Logo.

Plaintiffs claim rights in the purported mark “ibooks” for print and electronic books. (See
Am. Compl. ¶ 1.) The alleged mark appears on the copyright and the title page of Plaintiffs’
books, including Plaintiffs’ electronic books. (Mazzello Dec., Ex. 14 (Colby Dep. 218:19-219:6;
220:23-221:4).) Plaintiffs’ “ibooks” logo also purportedly appears inside all of Plaintiffs’

10

“ibooks” books.3 (See Mazzello Dec., Ex. 15 (Colby 30(b)(6) Dep. 319:19-22.); Mazzello Dec.,
Ex. 14 (Colby Dep. 219:16-20.).) Significantly, as shown below, Plaintiffs’ “ibooks” logo looks
nothing like Apple’s “iBooks” logo, which uses an open book on a wood background.

(Mazzello Dec., Exs. 17, 18.) Dr. McDonald, however, improperly failed to include the “ibooks”
logo in her survey. See Am. Footwear, 609 F.2d at 660 n. 4; THOIP, 690 F. Supp. 2d at 239.
2.

The McDonald Surveys Did Not Include Any Information About The
Copyright Or Title Page, Or The Appearance Of Plaintiffs’ Imprint.

Although respondents were told to “envision” a page in a digital book that “contains
information about the book—such as the date of publication, the publisher, the Library of
Congress number, etc.”—they were never shown pages from one of Plaintiffs’ books. Thus,
respondents were deprived of key information about the source of the book. For example, Dr.
McDonald admitted that Plaintiffs’ books Venus Prime and Dawn of Amber contained such
source identifying information as (1) the publisher’s website; (2) the name of the author; (3) the
name of the copyright owner; and (4) the Plaintiffs’ New York location—a long way from
Apple’s Cupertino, California office (see Mazzello Dec., Ex. 13 (McDonald Dep. 175:16182:11.)). None of this information, however, was included in her surveys.
Dr. McDonald contends that her surveys did not replicate book buying, but, instead,
recreated “the moment when a consumer becomes aware that there is something in a book that
identifies it . . . as iBooks.” (Id. (McDonald Dep. 139:25-140:8.).) According to Dr. McDonald,
this could occur upon the first viewing, or, “based on the appreciation of the consumer for the
3

In some instances, when a reader views one of Plaintiffs’ “ibooks” books on an e-reader device, the lightbulb logo
does not appear inside the electronic book, although it is “intended” to be there, due to the inability of the reader to
“pick[] up the jpeg that the bulb is in.” (Mazzello Dec., Ex. 14 (Colby Dep. 221:5-17.).)

11

book, and that kind of revisiting . . . .” (Id. (McDonald Dep. 140:11-15.).) As shown in the
excerpt from Plaintiffs’ book Voodoo Moon Trilogy, it would be nearly impossible to ignore the
numerous indicia of source that appear within Plaintiffs’ books:

The idea that an individual who was interested enough in the book to revisit the title page
to learn the publisher’s name would simultaneously ignore the plethora of source identifying
information—including the website address where additional information about the publisher
could be readily obtained—is more than implausible.

12

Thus, Dr. McDonald’s contrived scenario fails to replicate any real world situation,
whether it be pre-sale, point-of-sale, or, post-sale as there is no situation in which Plaintiffs’
“ibooks” imprint would be seen without other indicia of the book’s source.4 See Juicy, 2006 WL
1012939, *26 (holding that survey that showed respondents words on a card devoid of context
did not test point of sale or post-sale, confusion and finding no likelihood of confusion).
Further, the “ibooks” imprint is typically depicted in a specific font within Plaintiffs’
books. Mazzello Dec., Exs. 15 and 18 (Colby 30(b)(6) Dep. 319:19-320:5)). Dr. McDonald’s
survey prompt, however, provided no visual guidance regarding the appearance, size, or font of
the word “iBooks” or “ibooks.” C.f. Kargo, 2007 WL 2258688, at *11-12 (excluding survey
that, inter alia, showed an advertisement that depicted the mark larger than it appears in
advertisements seen by consumers.) In fact, the words “iBooks” and “ibooks” were written in
the same font and size as every other word in the survey prompt. (See Mazzello Dec., Exs. 2021.) Thus, as Dr. McDonald’s failure to replicate market conditions deprived respondents of
considerable contextual information, her surveys are not reliable and should be excluded. See
Juicy, 2006 WL 1012939, at *25; WE, 218 F. Supp. 2d at 474.
C.

The McDonald Surveys Placed Undue Emphasis on the iBooks Imprint

By telling respondents to envision the word iBooks in the abstract rather than in the
context of one of Plaintiffs’ books, Dr. McDonald placed undue emphasis on Plaintiffs’ imprint.
As Dr. McDonald admitted, “The first thing that consumers do when they pick up a book is
probably not look at the publisher.” (Mazzello Dec., Ex. 13 (McDonald Dep. 141:25-142:3.).)
Thus, Dr. McDonald designed surveys intended to replicate the moment when a consumer
becomes aware of the source identifying information within a book, i.e. notice the name of the
4

In addition, Plaintiffs’ electronic books are sold under the “ipicturebooks entity.” (Mazzello Dec., Ex. 19; see, e.g.
Exs. 5-8.) Nonetheless, Dr. McDonald did not include the source identifier “ipicturebooks” in her survey.

13

publisher, because such a moment “can occur . . . .” (Id. (McDonald Dep. 140:5-19 (emphasis
added).).) Admittedly, Dr. McDonald’s surveys “do[n’t] tell you how often that occurs,” nor
does she have third party research showing how often consumers notice the name of the
publisher. (Id. (McDonald Dep. 140:9-10.).) Thus, by artificially focusing respondents’
attention on the word “iBooks” or “ibooks,” which may easily go unnoticed in the real world, Dr.
McDonald overemphasized the role the “ibooks” imprint plays on an actual copyright or title
page. Cf. Vista Food Exch., Inc. v. Vistar Corp., 2005 WL 2371958, at *5 (holding that a survey
that depicted the plaintiff’s trademark as VISTA, when it was depicted as VISTA FOOD
EXCHANGE, INC. in the marketplace, “skewed results” because it improperly focused
respondent’s attention on the word VISTA, thus overemphasizing the role the word VISTA
played in the marketplace); Inc. Publ’g, 616 F. Supp. at 393 (holding telephone survey had no
probative value because it emphasized the aural elements of the mark, depriving respondents of
visual and cerebral stimuli and fragmenting the mark from its appearance and contents).
IV.

THE MCDONALD SURVEYS FAILED TO ASK THE GENERALLY
ACCEPTED QUESTION FOR TESTING CONFUSION.
“When unclear questions are included in a survey, they may threaten the validity of the

survey by systematically distorting responses if respondents are misled in a particular direction,
or by inflating random error if respondents guess because they do not understand the question.”
Reference Manual, at 388. “If the crucial question is sufficiently ambiguous or unclear, it may
be the basis for rejecting the survey.” Id.; see also Malletier, 525 F. Supp. 2d at 604 (excluding
survey where among other things the survey question was “highly ambiguous”).
The standard question for evaluating likelihood of confusion asks who or what company
makes or puts out the product. See McCarthy § 32.174. This question is derived from the “goldstandard” of surveys, the Ever-Ready survey, which was used in Union Carbide Corp. v. Ever14

Ready, Inc., 531 F.2d 366 (7th Cir. 1976), and has since become the standard methodology for
confusion surveys. Swann, Likelihood of Confusion, in Trademark and Deceptive Advertising
Surveys, at 53; see also McCarthy § 32.174. Here, however, Dr. McDonald chose not to follow
this typical question format, instead asking “what company or companies would you think had
made the book available?” (McDonald Rep. at 11 (emphasis added.).)
As Plaintiffs’ own expert Dr. Jacoby stated, this question cannot be used to assess
confusion here because it is “too ambiguous [and] just too broad.” (Mazzello Dec., Ex. 12
(Jacoby Dep. 247:7-22.).) As he explained, “[m]any people make the book available,” including
“the distributor [and] the retailer.” (Id. (Jacoby Dep. 247:7-248:7.).) Indeed, “Apple” is an
appropriate response to this question even for respondents who are not confused given that Apple
makes third party books available to consumers through Apple’s iBookstore and on its iPhone
and iPad devices. (See Mazzello Dec., Ex. 13 (McDonald Dep. 281:14-21.).) Apple also is
generally known for providing its iTunes service, an online retail store where third party content
may be downloaded, and thus even consumers who are not aware of Apple’s iBookstore would
be likely to accurately think of Apple in response to this question.
In fact, Dr. McDonald herself admits that the company that makes a book available “isn’t
necessarily the same as the company that prints the book [or] conceives of the book [or] edits the
book.” (Id. (McDonald Dep. 281:3-13.).) Instead, “available” in this context “could be
anybody.” (Id. (McDonald Dep. 280:13-16.).) Notably, Dr. McDonald admitted that only four
respondents named “the publisher” as the company making the book available (McDonald Rep.
at 15), suggesting that Dr. McDonald’s question was so vague and broad that it led respondents
away from thinking of the source of the book (or their affiliates or those who sponsored the
book) and toward thinking of companies that, like Apple, sell or distribute content created by

15

others. As Dr. McDonald’s question was designed to include “anybody” who had a hand in
making the book available — even someone who sold the book to consumers — this question is
incapable of assessing probative information and exclusion is warranted for this reason alone.
V.

THE MCDONALD SURVEYS FAILED TO USE A PROPER CONTROL
“‘[T]o prove actual confusion, the confusion must stem from the mark in question’ . . . .”

Malletier, 525 F. Supp. 2d at 596 (quoting General Motors Corp. v. Lanard Toys, Inc., 468 F.3d
405, 414 (6th Cir. 2006)). Thus, “[a] survey designed to estimate likelihood of confusion must
include a proper control.” THOIP, 690 F. Supp. 2d at 240. “Proper controls will approximate
‘background noise,’ the amount of confusion existing due to reasons unrelated to similarities”
between the marks at issue. Cumberland Packing Corp. v. Monsanto Co., 32 F. Supp. 2d 561,
574-75 (E.D.N.Y. 1999); see also Jacob Jacoby, Experimental Design and the Selection of
Controls in Trademark and Deceptive Advertising Surveys, Trademark Rep., Vol. 92, at 890 n. 2
(2002) (“[I]t is helpful to think of a control as a comparison used to rule out plausible alternative
explanations for the observed effect.”). “Without a proper control, there is no benchmark for
determining whether a likelihood of confusion estimate is significant or merely reflects flaws in
the survey methodology.” THOIP, 690 F. Supp. 2d at 240.
Here, Plaintiffs do not claim exclusive rights in the letter “i” standing alone, nor do they
claim rights in the word “books.” (See Am. Compl. ¶ 1.) They only claim rights in the
combination of “i” plus “books”—i.e., “ibooks.” (See id.) Further, Dr. McDonald admits that
the “i” prefix is a source indicator for Apple, as Apple owns a host of i-formative marks such as
iPad, iTunes and iPod. (See McDonald Rep. at 18.) Thus, even Dr. Jacoby agreed that a proper
control should use the letter i:
Q. . . . in light of the fact that for confusion to be actionable, it must be shown to be found
to be caused by the allegedly infringing action, not by something else, as we discussed,
would you agree that it would be important to use a control with "i" with something else
16

in order to filter out the people who might say Apple just because they think, oh, "i"
means Apple?
A. Yes.
(Mazzello Dec., Ex. 12 (Jacoby Dep. 169:11-22.).) 5
Dr. McDonald, however, used “eBooks” as her control and thus failed to account for
background noise associated with the letter “i.” Indeed, Dr. McDonald admits that 83% of the
respondents who thought that Apple had made the “iBooks” book available did so because the
“presence of the ‘i’ is associated with Apple.”6 (McDonald Rep. at 17.) Any confusion between
the “ibooks” imprint and the letter “i” is not actionable, however, as Plaintiffs have no claim to
exclusive rights in the letter “i” standing alone. A control should have been chosen to “net out
confusion” based on this irrelevant variable.7 Cumberland, 32 F. Supp. 2d at 575 (control was
improper because it did not account for confusion due to characteristics of the product other than
the claimed trade dress). The failure to do so warrants exclusion of the surveys. See THOIP,
690 F. Supp. 2d at 240-41 (excluding survey where, inter alia, control underestimated noise).

5

After a break in the deposition, during which Dr. Jacoby admittedly spoke to counsel about this aspect of his
testimony (See Mazzello Dec., Ex. 12 (Jacoby Dep. 266:7-15.)), he tried to retract this statement. Given the
apparent intervention of counsel during the break, his first answer is more credible.
6

In the Sur-Rebuttal Report, she stated that “[t]he vast majority (90%) noted that the presence of ‘i’ is associated
with Apple and Apple products.” (Sur-Rebuttal Rep. at 5.)

7

The control, “eBooks” is also improper because it is the generic term for electronic books and therefore cannot be a
“plausible member[] of the set of products at issue,” in this case, an imprint for a book. Shari Seidman Diamond,
Control Foundations: Rationale and Approaches, in Trademark and Deceptive Advertising Surveys, at 214 (2012)
In fact, Dr. Jacoby agreed that ebooks was an improper control for precisely the reasons identified above:
Q. Is it fair to say that you would not have used "ebooks" as a control, because that would not control for
the "i" and was a generic?
A. Yes.
(Mazzello Dec., Ex. 12 (Jacoby Dep. 176:20-24.).) Again, after speaking to counsel, Dr. Jacoby tried to retract this
statement. (Id. (Jacoby Dep. 266:7-15.).)

17

VI.

DR. MCDONALD’S SURVEYS SHOULD BE EXCLUDED AS WELL BECAUSE
HER METHODOLOGY SUFFERED FROM ADDITIONAL FATAL FLAWS.
Although each of the flaws above, standing alone, are a sufficient basis to exclude the

McDonald surveys, the surveys suffer from other flaws that also warrant exclusion.
A.

The McDonald Surveys Primed Respondents To Think Of Apple.

In the McDonald surveys, the screener asked respondents whether they “had downloaded
a digital book to a reading device of any kind, including a smartphone.” (McDonald Rep. at 9
(emphasis added).) Fifteen seconds later, they were asked if “you see the word ‘iBooks’ what
company or companies would you think had made the book available?” (Mazzello Dec., Ex. 13
(McDonald Dep. 259:2-6.).) As Apple is extremely well-known for making the iPhone
smartphone — and, in fact, has 48.1% of the smartphone market share in the United States (see
Soaring iPhone 5 sales in US knock Android into second place, Kantar World Panel (Nov. 11,
2012), http://www.kantarworldpanel.com/global/News/Soaring-iPhone-5-sales-in-US-knockAndroid-into-second-place) — referring to a “smartphone” in the screener question primed
respondents to think of Apple.8 Further, reference to a “smartphone made no sense because the
“device” could include any type of device on which books can be read. In fact, reading on a
smartphone seems much less common than reading on a larger device. This gratuitous reference
to smartphones was undoubtedly included just to prime respondents to think of Apple. Because
this sequence of questions likely inflated the rate of confusion, this is an independent basis for
excluding the survey. See Juicy, 2006 WL 1012939, at * 26 (survey had no value where, inter
alia, introductory question referenced clothing suggesting the survey was connected to clothing).

8

Moreover, respondents could, and some in fact did, take the McDonald surveys on an Apple device, including
Apple’s iPhone or iPad. (See Mazzello Dec., Ex. 13 (McDonald Dep. 259:18-260:3.); Sur-Rebuttal Rep., Appendix
B, Respondent 60001085.) This could likely have also led respondents to think of Apple when taking the survey.

18

B.

The McDonald Surveys Did Not Test The Proper Universe.

Contrary to the established procedure for likelihood of confusion surveys, in which one
examines the response of potential future purchasers of the product in issue, Dr. McDonald only
screened for individuals who had downloaded a digital book in the past.9 (McDonald Rep. at
11.) However, “[i]t is well-established that only potential future purchasers, not past purchasers,
are relevant to a confusion study.” Malletier, 525 F. Supp. at 604. Thus, Dr. McDonald’s
universe was improper for two reasons. First, it included those who possibly had never
purchased an electronic book because many electronic books can be downloaded for free.
Second, it included those who possibly had no interest in purchasing a book in the future because
she only screened for past behavior. Therefore, her universe was improper.10 See Universal City
Studios, Inc. v. Nintendo Co., Ltd., 746 F.2d 112, 118 (2d Cir. 1984) (excluding survey that, inter
alia, included past rather than future purchasers); Am. Footwear, 609 F.2d at 660 n. 4 (affirming
rejection of a survey that did not include people with a future purchasing interest); Malletier, 525
F. Supp. 2d at 604 (excluding survey that, inter alia, included past rather than future purchasers).
In addition, Dr. McDonald’s surveys were materially under-inclusive because she “did
not include people who read print books only.” (Mazzello Dec., Ex. 13 (McDonald Dep. 221:22222:10.).) This universe is improper because

of Plaintiffs’ sales have been for print

books.11 (Mazzello Dec. ¶¶ 3-9.) Further, even assuming that electronic books have made up
9

Dr. McDonald also required that respondents “indicate[] they read books on a regular or fairly regular basis . . . .”
McDonald Rep. at 11.) This requirement also is improper. Rather than use an objective measurement, Dr.
McDonald selected a “totally subjective” term. (Mazzello Dec., Ex. 13 (McDonald Dep. 231:19-21.).) Thus, her
surveys did not include individuals who read on a regular basis, but only those who perceive themselves to be
regular readers. It also excluded people who purchase, but do not read books, such as those who buy books as gifts.

10

After reading Dr. Nowlis’ criticism, Dr. McDonald testified that the invitation for her survey was only sent to
Research Now panelists who had “bought or shopped for a book.” (Mazzello Dec., Ex. 13 (McDonald Dep. 209:1318.).) There, however, were no questions in the screener regarding purchasing behavior, either past or future, and
Plaintiffs have not produced either the invitation or documents regarding the selection of panelists.
11

This figure includes books bearing either the “ibooks” or “ipicturebooks” imprint. (Mazzello Dec., Ex. 15 (Colby
30(b)(6) Dep. 178:23-179:7.).)

19

1.83% of Plaintiffs’ sales is generous, as most of those sales were accounted for by a single book
offered on a CD-ROM in 2001 under the name “ipicturebooks.” (Id. (Colby 30(b)(6) Dep.
184:11-186:8.).) Even Dr. Jacoby admitted that “it’s important to include potential readers of
print books” in the universe in this case. (Mazzello Dec., Ex. 12 (Jacoby Dep. 108:7-8.).) In
addition, in narrowing her universe to people who had downloaded an electronic book, Dr.
McDonald herself admitted that she selected a universe of consumers who are “far more aware
of [Apple’s] iBooks as a brand” (Mazzello Dec., Ex. 13 (McDonald Dep. 224:7-8.).) and
therefore more likely to name Apple in response to her survey questions. Thus, while a survey
universe does not have to be perfect, and courts have found that surveys that exclude a small
portion of the relevant universe are probative, in this situation, where sales are almost
exclusively for print books, Dr. McDonald’s universe of those who previously downloaded
electronic books is impermissibly under-inclusive. See Conopco, Inc. v. Cosmair, Inc., 49 F.
Supp. 2d 242, 253 (S.D.N.Y. 1999) (survey did not establish confusion because it excluded a
“substantial portion” of the relevant universe); William G. Barber, The Universe, in Trademark
and Deceptive Advertising Surveys, 43-4 (2012) (“where the excluded consumers make up a
sizeable proportion of the relevant universe, courts tend to view the survey with skepticism”).
C.

The McDonald Surveys Did Not Properly Screen Respondents.

It is standard survey practice to screen out atypical respondents, such as those who work
in advertising, those who work in the industry at issue, or those who are “professional
respondents” or recently had taken a survey in a related industry, as they have specialized
knowledge that may bias the survey. See Barber, at 49; (Mazzello Dec., Ex. 12 (Jacoby Dep.
152:2-11; 321:9-324:15.).) Here, not only did Dr. McDonald fail to exclude these people, she
drew 70% of her respondents from Research Now, an online panel of people who “have agreed

20

to receive web invitations to participate in surveys.” (Id. (McDonald Dep. 204:3-4; 209:12-13;
244:8-23; 249:3-10.).) Such participants are the very definition of “professional respondents.”
Compounding this error, Dr. McDonald made no effort to determine the demographic
make-up of ebook purchasers because, contrary to standard practice, she believed this
information was “irrelevant.” (Id. (McDonald Dep. 236:4-9.).) Thus, she did not ensure that the
demographic make-up of her survey sample was representative of the demographic make-up of
ebook purchasers. This error further undermines the reliability of the McDonald surveys. See
Mastercard, 2004 WL 326708, at *9-10 (excluding survey where, inter alia, there was no proof
the respondents were representative of all qualified respondents); Barber, at 48 (“[t]he sampling
frame should be designed to encompass a representative cross-section of the entire universe”).
D.

The McDonald Surveys Failed To Take Steps To Prevent Guessing.

A properly conducted survey should instruct respondents not to guess. See Malletier, 525
F. Supp. 2d at 568-69 (excluding survey that did not discourage guessing among other things);
Cumberland, 32 F. Supp. 2d at 572 (surveys flawed and unpersuasive for failing to instruct
against guessing, among other things). The McDonald surveys did not include such an
instruction. In fact, Dr. McDonald, consistent with her disregard for time-tested, court-approved
approaches to survey design, “actively disagrees with” the practice of instructing respondents not
to guess and stated that “not withstanding whatever jurists may say, . . . I believe that this is not
an appropriate injunction.” (Mazzello Dec., Ex. 13 (McDonald Dep. 263:23-264:5; 272:20-23.).)
This resulted in a considerable amount of guessing as indicated by the verbatim responses, which
include numerous statements such as “I would guess Apple . . .” (McDonald Rep., App. D,
Respondent 100001749) and “This is a total guess but I would say Apple . . .” (Sur-Rebuttal
Rep., App. B, Respondent 60003016) — all of which Dr. McDonald counted as confused.
Because this guessing inflated confusion and, the surveys are unreliable.
21

E.

Dr. McDonald Improperly Calculated The Rate Of Confusion.

“‘In order to prove actual confusion, the confusion must stem from the mark in question’
. . ..” Malletier, 525 F. Supp. 2d at 596 (quoting General Motors Corp., 468 F.3d at 414). Thus,
when calculating the confusion rate, “[c]onfusion caused by stimuli irrelevant to the trademark at
issue should be disregarded.” Id. Here, Dr. McDonald counted as confused individuals who
identified Apple or iTunes because of the presence of the letter “i” alone. (McDonald Rep. at
17.) Because Plaintiffs do not claim rights in the letter “i” alone, respondents who identified
Apple solely because of this letter should not be counted as confused. See THOIP v. Walt
Disney Co., 788 F. Supp. 2d 168, 182 (S.D.N.Y. 2011) (excluding survey where calculation
improperly included as confused respondents whose answers were “not based on the purportedly
protectable elements of THOIP’s mark”); Malletier, 525 F. Supp. at 596-97 (expert improperly
counted as confused respondents who referred to similarities in non-protectable elements of the
trade dress). In other words, Apple has worked hard to develop a family of i-formative marks
that people properly associate with Apple. Plaintiffs should not be able to hijack Apple’s
permissible use of “i,” and essentially use Apple’s unrelated fame to make its case. As discussed
by Dr. Nowlis, when these improperly counted respondents are eliminated from the first survey,
the number of confused respondents drops dramatically from 53% as calculated by Dr.
McDonald to 13%. (See Mazzello Dec., Ex. 11 ¶ 64.).)
Nor is that Dr. McDonald’s only miscalculation. According to Dr. Nowlis, in the first
survey, another 16 respondents who were counted as confused named Apple, but offered no
explanation or said they did so for reasons other than the word “ibooks.” (Id., Ex. ¶ 65.).) These
16 respondents also should not be counted as confused because their responses were not based

22

on the allegedly infringing mark.12 See Mastercard, 2004 WL 326708, at *10 n.12 (holding that
the confusion rate was calculated improperly because it counted as confused those who based
their answer on factors other than the trademark in issue); Cumberland, 32 F. Supp. 2d at 575
(same). After eliminating these 16 respondents and subtracting the control rate of confusion, the
final confusion rate for the initial McDonald survey is a mere 5.2%.13 (See Mazzello Dec., Ex.
11 ¶ 66-7.) This low confusion rate cannot support a finding of reverse confusion. See
McCarthy § 32:189 (confusion rates below 10% can become evidence that confusion is not
likely). Moreover, the failure to accurately calculate the confusion rate raises further doubts
about the surveys’ reliability and relevance. See THOIP, 788 F. Supp. 2d at 182); Malletier, 525
F. Supp. at 596-97 (same); Mastercard, 2004 WL 326708, at *10 n.12 (same).
F.

The McDonald Surveys Were Not Validated.

It is standard procedure to validate a litigation survey in order to ensure that those who
took the survey were qualified to do so and in fact did so. See Reference Manual, at 411-12.
This procedure makes even more sense in Internet surveys, where the identity of the respondent
is unknown. Without validation, there is no way of knowing if the person who took the Internet
survey was, in fact, the person who signed up for the panel and received the survey invitation.
Dr. McDonald, however, did not validate her surveys, further departing from standard practice.
(Mazzello Dec., Ex. 13 (McDonald Dep. 320:9-10.).)

12

Notably, Dr. McDonald admitted that she counted as confused all respondents who named Apple (and no other
company) regardless of the explanation offered. (Mazzello Dec., Ex. 13 (McDonald Dep. 308:24-310:16.).)

13

The Sur-Rebuttal survey suffered from similar errors in calculating the confusion rate. Although Dr. McDonald
claimed the survey had a confusion rate of 53% (Sur-Rebuttal Rep. at 3), she once again counted as confused
respondents who (1) named Apple because of the presence of the “i” alone, (2) failed to offer any reason for naming
Apple, or (3) cited reasons other than the mark at issue for naming Apple. After applying Dr. Nowlis’ methodology
to the sur-rebuttal calculation, the final confusion rate drops to 3.8%. (Mazzello Dec. ¶¶ 15-18.)

23

G.

Dr. McDonald Improperly Speculates About The Future.

“[E]xpert testimony should be excluded if it is ‘speculative or conjectural’ . . . .” Lippe v.
Bairnco Corp., 288 B.R. 678, 686 (S.D.N.Y. 2003) (citation omitted) aff'd, 99 F. App'x 274 (2d
Cir. 2004). Not only did Dr. McDonald opine on the present using her flawed initial survey, but
she opined on the future, stating that she “can easily offer the opinion that the level of source
confusion would be even larger in a study conducted several years from now.” (McDonald Rep.
at 18.) She also claims that her initial survey “confirms” that “Apple’s mark creates . . .
potentially greater confusion in the future.” (McDonald Rep. at 1.) By speculating about the
future, she yet again departed from the principles of scientific research, confirming her antiApple bias and apparent desire to advocate Plaintiffs’ position as much as possible.
VII.

DR. MCDONALD IMPROPERLY OPINES ON PLAINTIFFS’ AND APPLE’S
BRAND DEVELOPMENT
Dr. McDonald’s first report included a preamble to the discussion of her initial survey in

which she generally discussed Plaintiffs’ use of the “ibooks” imprint and Apple’s alleged brand
development strategy. (See McDonald Rep. at 1-7.) At her deposition, Dr. McDonald revealed
that she considered this her marketing opinion and implied that it was somehow separate from
her survey work. (See Mazzello Dec., Ex. 13 (McDonald Dep. 65:20-66:14.).) Were this true,
this portion of her report should be stricken because, as Dr. Nowlis explained in his deposition,
Dr. McDonald cites to no evidentiary support for any of her propositions. (Mazzello Dec., Ex.
16, (Nowlis Dep. 101:20-102:24; 121:24-122:7.).) In fact, she admits that she “did no
independent research . . . not even web surfing” (Mazzello Dec., Ex. 13 (McDonald Dep. 74:1975:3.)), nor did she review any deposition testimony or Apple’s trademark clearance searches for
the iBooks mark (id. (McDonald Dep. 48:15-49:13.)). Instead, she admitted that she relied on
the allegations in the Amended Complaint with “no further investigation.” (Id. (McDonald Dep.
24

70:23-71:13; 81:21-82:25.).) In light of the fact that she admits that she is not a publishing
expert or computer software expert (id. (McDonald Dep. 18:9-21.)), and thus cannot rely on her
experience in these industries, it is particularly important that she cite some basis for her opinion.
Thus, Dr. McDonald’s opinions in this section of her report are wildly speculative and
wholly unsupported by any factual data or methodology. She relies only on “[her] own say-so or
on unfair and one-sided interpretations of the available [information],” and therefore her
marketing opinions are unreliable and must be excluded less they prejudice any potential jury.
Lippe, 288 B.R. at 701; see also Weiner v. Snapple Beverage Corp., 07 CIV. 8742 DLC, 2010
WL 3119452, at *8-10 (S.D.N.Y. Aug. 5, 2010) (excluding expert testimony that was based only
on the complaint and two internal marketing documents as unreliable).
Date: December 21, 2012

Respectfully submitted,
s/ Dale M. Cendali
Dale M. Cendali
Claudia Ray
Bonnie L. Jarrett
KIRKLAND & ELLIS LLP
601 Lexington Avenue
New York, New York 10022
Tel: [phone]
Fax: [phone]

Perry J. Viscounty
LATHAM & WATKINS LLP
140 Scott Drive
Menlo Park, CA 94025
Tel: [phone]
Fax: [phone]

Jennifer L. Barry
LATHAM & WATKINS LLP
600 West Broadway, Suite 1800
San Diego, CA 92101-3375
Tel: [phone]
Fax: [phone]
ATTORNEYS FOR DEFENDANT

25