United States District Court, E.D. Michigan, Southern Division
VISTEON GLOBAL TECHNOLOGIES, INC. AND VISTEON TECHNOLOGIES, LLC, Plaintiffs,
GARMIN INTERNATIONAL, INC., Defendant.
R. Grand United States Magistrate Judge
AND ORDER (1) GRANTING GARMIN'S MOTION TO EXCLUDE DR.
JOEL STECKEL'S CONJOINT SURVEY AND CONCLUSIONS DERIVED
THEREFROM (ECF NO. 259) AND (2) GRANTING GARMIN'S MOTION
TO EXCLUDE PETER SMITH'S REASONABLE ROYALTY CALCULATIONS
AND OPINIONS REGARDING DAMAGES (ECF NO. 260); (3) GRANTING
GARMIN'S MOTION TO STRIKE DR. STECKEL'S DECLARATION
(ECF NO. 318); (4) GRANTING GARMIN'S MOTION TO STRIKE
PORTIONS OF PETER SMITH'S EXPERT REPORT (ECF NO. 342);
AND (5) GRANTING VISTEON'S DAUBERT MOTION TO EXCLUDE JOHN
LAVRAKAS FROM OFFERING ANY TESTIMONY REBUTTING SURVEY (ECF
D. BORMAN UNITED STATES DISTRICT JUDGE
patent infringement action, Defendant Garmin International,
Inc.'s (“Garmin”) has moved to exclude the
expert opinions of Plaintiffs Visteon Global Technologies,
Inc. and Visteon Technologies, LLC's
(“Visteon”) damages experts Dr. Joel Steckel and
Mr. Peter Smith. (ECF Nos. 259, 260.) Visteon has filed
Responses to both motions (ECF Nos. 280, 281) and Garmin has
filed Replies (ECF Nos. 319, 320). The Court held a
Daubert hearing on July 13, 2016, and heard
testimony from both Dr. Steckel and Mr. Smith. The parties
filed post-hearing briefs on August 12, 2016. (ECF Nos. 376,
an action for patent infringement. In the claims that remain
for trial, Visteon Global contends that Garmin infringes,
either directly or indirectly, U.S. Patent No. 5, 544, 060
(“the ‘060 patent”), U.S. Patent No. 5,
654, 892 (“the ‘892 patent”) and U.S.
Patent No. 5, 832, 408 (“the ‘408 patent”).
In general, the '060 patent is directed to a method of
navigating a vehicle whereby a user can generate an optimal
path and then switch to an alternate navigation path before
beginning on the optimal path. Visteon contends that Garmin
directly infringes Claims 3, 4, and 6 and indirectly
infringes Claim 3 of the ‘060 patent. In general, the
‘892 patent is directed to a method for assisting the
navigation of a vehicle whereby a complex arrow icon is
generated and displayed to the driver at a predetermined time
or distance before the driver reaches a particular maneuver.
Visteon contends that Garmin directly and indirectly
infringes Claim 8 of the ‘892 patent. In general, the
‘408 patent is directed to a navigation system which
allows the user to search for a destination either from a
list of categories or from an alphanumeric search. Visteon
contends that Garmin directly and indirectly infringes Claims
4 and 5 of the ‘408 patent.
contends that a variety of Garmin navigation products
(“the accused products”) infringe one or more the
asserted claims literally and under the doctrine of
equivalents. Visteon contends that Garmin directly infringes,
actively induces infringement and/or contributorily infringes
one or more of the asserted claims of the asserted patents,
and that Garmin's infringement is and has been willful.
Visteon seeks damages in the form of a reasonable royalty for
Garmin's alleged infringement. Visteon also seeks
prejudgment interest, enhanced damages, attorneys' fees,
and costs, including without limitation any fees and costs
associated with participating in multiple ex parte
proceedings at the United States Patent and Trademark Office
(“USPTO”) as initiated by Garmin in this action.
Garmin denies that it directly, indirectly or contributorily
infringed or induced infringement of any of the patents in
suit and affirmatively asserts a host of invalidity defenses.
Garmin seeks judgment in its favor and requests that it be
awarded its costs and reasonable attorneys' fees incurred
in defending against Visteon Complaint.
respect to Dr. Steckel, Garmin attacks the “scientific
adequacy” of Dr. Steckel's survey, challenging
aspects of the underlying design of the survey and the
analysis that Dr. Steckel employed in analyzing the data that
he collected. Garmin levels three principal attacks on the
reliability of Dr. Steckel's methods as applied in this
case and to his conclusions, which Garmin argues resulted in
inflated values of the accused features “far beyond any
semblance of real-world credibility.” Garmin argues
that the survey produced “economically irrational
results, ” attributable to at least the following
fundamental flaws: First, Garmin argues that Dr. Steckel did
not survey the actual patented functionalities but rather
tested features much broader than the accused features,
leading to an overvaluation of the patented features. Second,
Garmin argues that Dr. Steckel failed to include reasonable
distracter features in his product profiles, thus
inordinately focusing respondents on the patented features.
And third, Garmin argues that Visteon improperly extrapolates
from Dr. Steckel's “economic values” to
support their reasonable royalty damage calculation.
respect to Mr. Smith, Garmin levels three principal attacks
on his reasonable royalty calculation: (1) Mr. Smith's
reliance on Dr. Steckel's Economic Values fails to
apportion revenue to the individual allegedly infringing
features (as distinct from the multitude of non-patented
features in the accused devices); (2) Mr. Smith failed to
properly apply the Georgia-Pacific analysis; and (3)
Mr. Smith failed to explain the methodology he applied to
reach his final royalty calculation that was revised to
account for Dr. Steckel's own revision to his Economic
Dr. Steckel's Conjoint Study
survey expert, Dr. Joel Steckel, is a Professor of Marketing
at the Stern School of Business, New York University, where
he is the Vice Dean of Doctoral Education. Dr. Steckel
received his B.A. from Columbia University and his M.B.A.,
M.A., and Ph.D degrees from the Wharton School, University of
Pennsylvania. Dr. Steckel has visiting or permanent faculty
appointments at the Columbia University Graduate School of
Business, the Anderson Graduate School of Management,
U.C.L.A., the Yale University School of Management and the
University of Pennsylvania Wharton School. (ECF No. 268,
Garmin Index of Exhibits, Ex. E, 11/30/12 Revised Report of
Joel Steckel, Ph.D ¶ 1). (“11/30/12 Steckel
Report”). Dr. Steckel's fields of specialization
include marketing strategy, marketing research, and consumer
decision-making. Dr. Steckel has taught M.B.A. students
about, written textbook chapters on, and lectured executives
on conjoint analysis. Id. ¶ 3. At the
Daubert hearing, Dr. Steckel first testified that
this was the first time he had performed a conjoint analysis
in a patent matter, but later at the hearing recalled having
done a conjoint study in one other patent case that never
went to trial. (ECF No. 386, Sealed Transcript of July 13,
2016 Daubert Hearing at 37:24-38:13.)
Steckel conducted a choice-based conjoint (“CBC”)
consumer survey on behalf of Visteon to attempt to determine
the value that consumers place on the individual accused
patented features. Conjoint analysis “is based on the
notion that consumers will “consider
jointly” the attributes or characteristics of a
product when making their purchasing decisions.”
Steckel Report ¶ 36 (emphasis in original). V. Seenu
Srinivasan, a tenured Professor Emeritus at the Stanford
Graduate School of Business, and an acknowledged
“father of conjoint analysis, ” explained the
survey-based research approach as reported in TV
Interactive Data Corp. v. Sony Corp., 929 F.Supp.2d 1006
(N.D. Cal. 2013):
Professor Srinivasan describes conjoint analysis as a type of
survey or market research, which, at the most general level,
conceptualizes products as bundles of attributes, treating
price as an attribute. [Declaration of Professor V. Seenu
Srinivasan]. ¶ 3. Conjoint analysis uses customer
surveys to determine “values” for each attribute.
By choosing among multiple bundles of attributes, survey
participants make implicit tradeoffs one would make in
real-world purchasing decisions. Id. “For
example, conjoint analysis offers respondents hypothetical
products in several combinations, some of which might contain
feature 1 (but not feature 2), and some of which might
contain feature 2 (but not feature 1).” Id.
Professor Srinivasan explains that by comparing
respondents' choices when presented with different
features, one can estimate the quantitative values of
specific features. Id. According to Professor
Srinivasan, studies have validated that this implicit
tradeoff is more reliable than asking consumers directly what
they would pay for a specific feature. Id.
929 F.Supp.2d at 1020.
Steckel's conjoint study was a web-based study that
respondents took on their desktop or laptop computers.
Id. ¶ 42. Dr. Steckel worked with a
professional survey firm, Applied Marketing Science
(“AMS”), that implemented the survey under Dr.
Steckel's direction. Id. ¶ 10. A conjoint
survey typically asks study participants to compare a number
of product profiles to determine a relative value that the
consumer attaches to a particular, individual feature or
attribute. “By capitalizing on respondents' joint
consideration of the various attributes for specified
products, the research is then able to analyze the response
data to evaluate consumer preferences and relative valuations
for individual product attributes, combinations of product
attributes, and particular products.” Id.
¶ 36. A respondent in a conjoint consumer survey is
asked to choose from among a “choice set” of
products, with each choice set described by profiles of
bundled attributes. Id. ¶ 37. For example, a
conjoint exercise involving bottled water might contain
attributes such as flavor, label color, carbonation and
bottle shape. Each attribute would be further defined by a
set of characteristics known as “levels, ” for
example the flavor attribute would have multiple mutually
exclusive possible levels, such as lemon, lime, and orange.
Steckel selected six attributes to define the GPS product
profiles shown to respondents in the conjoint survey: a.
Points of Interest (the ‘408 patent), b. Special
Designation Selection (the ‘375 patent), c. Preview and
Route Adjustment (the ‘060 patent), d. Turn Preview
Display (the ‘892 patent), e. Language Display and f.
Price. Id. ¶ 38. Attributes a, b, c, and d are
related to the patented features, attribute e is a
“distracter, ” and attribute f allows a
measurement of value associated with all of the other
attributes. Id. Each attribute, with the exception
of price, is defined by two levels. Id. ¶ 39. A
sample three-product profile choice task would look like
price as one of the attributes “allows for the
measurement of an economic value for each patented feature
relative to appropriate reference points.” Id.
¶ 40. Dr. Steckel utilized three device product profiles
per choice task and in each instance asked the respondent to
select the device that they would be most likely to purchase.
Id. ¶ 43. Respondents were shown a total of 16
choice tasks to complete the survey. Id.
on the respondent's selections from the choice sets, Dr.
Steckel applied a well established statistical analysis,
utilizing the multinomial logit model employing Hierarchical
Bayes (“HB”) estimates, to determine the separate
value, or “partworth, ” that consumers attach to
each product attribute. Id. ¶¶ 44-45.
Partworths reflect how much each attribute contributes to the
overall consumer preference for a product (here a feature)
and allow the researcher to assess the relative importance of
attributes by examining the ranges of partworths.
Id. ¶ 44. The partworths allow an assessment of
the importance of the features described by the attributes in
consumer purchasing decisions and allow an evaluation of the
willingness of the consumer to trade off features against
price. Id. ¶ 45. Dr. Steckel used the
partworths on price and the attributes to estimate the extent
of these tradeoffs for each of the patented features,
resulting in an “economic value” or
“EV” for the patented feature relative to the
closest allegedly non-infringing alternative. Id.
the respondents to the conjoint survey were recruited through
Survey Sampling International (“SSI”), a company
that pre-recruits potential respondents who have indicated a
willingness to participate in such market research surveys.
Id. at ¶ 18. SSI respondents were invited to
take the survey based on their profile and demographics.
Id. at ¶ 19. Several controls were in place to
ensure that only invited respondents could respond, that each
respondent could only complete the survey one time and that
the respondent's stated age and gender matched their
profile in the SSI database. Id. Forty respondents
to the conjoint survey were excluded because their age and
gender did not match the SSI database profile. Id.
¶ 49. A total of 737 invited respondents qualified to
participate and, out of those 737, 588 respondents completed
the conjoint survey between May 7 and 11, 2012. Id.
¶ 50. The initial results from the 588 respondents who
completed the survey were screened for basic unreliability,
such as completing the survey too quickly or too slowly or
with straight-line responses that demonstrate insufficient
attention to the survey. Id. ¶ 52. There were
546 respondents in the final sample set and they took an
average of 27.3 minutes to complete the survey. Id.
designing the survey, Dr. Steckel undertook two levels of
exploratory research to determine the efficacy of the design.
First, AMS interviewed 15 GPS users to determine an
appropriate vocabulary for the questionnaire based on the
words and phrases commonly used by consumers. In a second
round of exploratory research, AMS interviewed 25 GPS users
to ensure that the survey set forth the closest allegedly
non-infringing attribute alternative. Id.
¶55(a). In making the final selection of non-infringing
alternatives to include in the survey, Dr. Steckel was
informed by the results of his exploratory research and also
was assisted by counsel for Visteon. ECF No. 268, Garmin
Index of Exhibits, Ex. F, 11/20/12 Deposition of Dr. Joel
Steckel 138:25-139:7. None of the individuals who
participated in the exploratory research were asked to
complete the conjoint survey. Steckel Report ¶
55(a)(iii). Before going live with the conjoint survey, Dr.
Steckel conducted a web-based pretest of 10 GPS consumers who
were debriefed after they completed survey. Id.
¶ 55(b). The survey presented video depictions of the
attributes and levels, utilizing voice-over, animation and
accompanying text to enhance visual explanations.
Id. ¶ 55(c).
beginning each choice task, respondents were given
descriptions of each of the attributes and levels through
general text descriptions, video animations and charts
describing the various options that respondents would see in
the upcoming choice task. Steckel Report ¶ 56. Each
respondent was then given 16 separate choice tasks, each task
consisting of three different GPS devices with various
attributes and levels within those attributes, and were told
to select the device that they would be most likely to
purchase and to assume that they must purchase one, i.e.
there was not a “no-purchase” option. Steckel
Report ¶¶ 57, 58; Steckel Report Ex. H-19.
Respondents were told to assume that aside from the
attributes identified in the choice sets, the devices were
Steckel used the data generated by the respondents across all
of their choice tasks to estimate the partworths, i.e. the
value that respondents placed on each of the levels within
each attribute. Steckel Report ¶ 59. Dr. Steckel used
software developed by Sawtooth Software, Inc. to obtain
individual respondent-level HB estimates of the partworths.
The partworths generated by the HB method are measured in
units of consumer utility - if a consumer's utility for
attribute level A is greater than that of attribute level B,
then that consumer prefers attribute level A to attribute
level B, all else being equal. Id. In obtaining the
HB estimates, Dr. Steckel constrained the partworths on the
price attribute for each respondent to be monotonic, i.e. his
analysis imposed the reasonable economic assumption that
individuals prefer paying a lower price for a product than a
higher price for that same product. Id. ¶
Steckel's conjoint survey results indicated that the mean
partworth on the attribute level relating to the patented
feature was greater in the case of each of the patented
attributes than the partworth on the level associated with
the non-infringing alternative for that attribute. Steckel
Report ¶¶ 63, 64. Dr. Steckel used the
respondent-level partworths on price and the other attributes
to estimate the EV of the patented features by expressing in
“units of utility” as a percentage of a reference
price. Id. ¶ 66. Dr. Steckel performed this EV
calculation for each respondent for each of the patented
features and then computed a simple mean EV for the presence
of the patented feature versus its absence. This computation
yielded the following results: “Respondents indicated
that they are willing to pay, on average, $13.93 for the
presence of the Points of Interest Feature ('408 patent),
$14.48 for the Preview and Route Adjustment Feature ('060
patent), $11.86 for the Special Destination Selection Feature
('375 patent), and $10.41 for the Turn Preview Display
feature ('892 patent). Id. 68.
using the term “willingness to pay” in his
Report, Dr. Steckel emphasized that his economic values do
not represent the actual amounts that consumers would be
willing to pay for inclusion of the patented features in a
competitive market. Id. ¶ 69. Dr. Steckel
expressly stated that “it would be incorrect to suggest
that these four patented features alone are responsible for
$50.68 of the price of a GPS system.” Id. Dr.
Steckel confirmed this understanding of his economic values
at the Daubert hearing, testifying expressly that
“the Economic Values do not reflect what a consumer
would be willing to pay for the attribute in the marketplace
necessarily.” Daubert Tr. 24:22-23.) “Q:
Did you take any steps as part of your survey analysis to
determine the price of these features in the actual
marketplace? A: I did not.” Daubert Tr.
to the Court's inquiry as to why he elected to use only
one relatively unimportant distracter feature, i.e. language,
in his study, Dr. Steckel explained that it was unnecessary
for purposes of his study to determine the actual value of
the patented features to consumers in relation to the many
other non-patented features of the accused devices:
I don't need to create the real world, completely real
world trade-offs for my purposes, for the purposes of the
Study. If I were trying to project consumer choices, so how
market share might have been adjusted but for the
incorporation of the patent features, then [mulitple
distracters] would have been really important, then they
would have been critical. I believe in those cases if that
were my goal, Garmin's criticisms would be appropriate.
However, that was not my goal. I am just trying to create the
trade-offs between the particular features and money without
projecting to real world choices. So I think the short answer
to your question is I wasn't trying to project to real
world choices - I don't need to for purposes of the Study
as I described them.
Daubert Tr. 37:5-14 (alteration
added). Dr. Steckel does suggest that the economic
values do provide an indication that in general the patented
features are of significant value to consumers and that not
all GPS features are valued equally by consumers. Smith
Report ¶ 69.
pre and post-Daubert hearing briefing makes clear,
and it is undisputed, that Dr. Steckel's study did not
attempt to determine a real world price for the four patented
features, and did not endeavor to value any non-patented
features or to determine the value of the four patented
features relative to the multitude of non-patented features
in the accused devices: “Garmin correctly posits that
‘[c]onjoint surveys, when done properly, are useful for
determining the relative value of one
feature to other tested features.'” (ECF No. 280,
Visteon's Resp. to Garmin's Mot. to Exclude Smith 5
(“Smith Resp. Br.”) (Emphasis and alteration in
original). The only tested features in Dr. Steckel's
study, apart from the unimportant language distracter, were
the four patented features.
the only “value” expressed in Dr. Steckel's
economic values is the relative value of the four asserted
patented features to one another. Dr. Steckel's conjoint
results express nothing about the value of the four patented
features relative to other important features of the accused
devices. “Dr. Steckel's conjoint analysis was used
for one (and only one) purpose by Mr. Smith - to apportion
the value of the patented features relative to each
other.” (ECF No. 376, Visteon's
Post-Daubert Hearing Brief 1.) (Emphasis in
original). To emphasize this point, Visteon quotes the
following from paragraph 69 of Dr. Steckel's Report,
portions of which were quoted above:
It is important to recognize that these values do not
represent the actual amounts consumers would be willing pay
for the inclusion of the patented features in a competitive
market. It would be incorrect to suggest that these four
patented features alone are responsible for $50.68 of the
price of a GPS system. Price is primarily determined by three
factors: consumer value, producer costs, and competition. I
studied only one, consumer value.
Smith Resp. Br. 5 (emphasis in original).
discussed at length infra, it is Visteon's
burden in this case to tie its reasonable royalty damage
calculation to the actual incremental value of the patented
features in the real world marketplace, i.e. to establish the
incremental profit that Garmin could have expected to earn
from incorporating the patented technology in these four
patents into their PNDs at the time of a hypothetical
negotiation. Visteon concedes that Dr. Steckel did not
attempt to do this but claims that it has satisfied this
burden through additional expert evidence provided by Mr.
Peter Smith, whose opinion builds on Dr. Steckel's work:
[P]rice (or actual value apportioned to the patents-in-suit)
can be determined when the Economic Values are combined with
producer costs and competition. With this
in mind, Mr. Smith picked up where Dr. Steckel's analysis
left off, and provided the two other elements of the price
calculation via Georgia-Pacific analysis and
Garmin's target profit margin: producer costs and
competition. . . . Garmin's target profit margin directly
accounts for its royalty and hardware costs (producer
costs) and was derived by Mr. Smith as a first
multiplier to begin to transform the Economic Values from the
theoretical consumer values to the prices attributable to the
patents-in-suit. The next factor was derived from Mr.
Smith's Georgia-Pacific analysis which
considers, among other things, competition in the
industry. . . . Mr. Smith's analysis is solidly based on
a reliable methodology for attributing value to the patented
features, and thus complies with the Court's
apportionment requirement. Contrary to Garmin's assertion
that “Mr. Smith did not take any additional
steps” to apportion, Mr. Smith's analysis completes
the apportionment process started by Dr. Steckel to assign
value to each of the patented features.
Smith Resp. Br. 5-7 (emphasis in original).
Mr. Smith's Reasonable Royalty Calculations
a finding of infringement, “the court shall award the
claimant damages adequate to compensate for the infringement,
but in no event less than a reasonable royalty for the use
made of the invention by the infringer.” 35 U.S.C.
§ 284. “The most common method for determining a
reasonable royalty is the hypothetical negotiation approach,
which ‘attempts to ascertain the royalty upon which the
parties would have agreed had they successfully negotiated an
agreement just before infringement began.'”
VirnetX, Inc. v. Cisco Sys., Inc., 767 F.3d 1308,
1326 (Fed. Cir. 2014) (quoting Lucent Techs., Inc. v.
Gateway, Inc., 580 F.3d 1301, 1324 (Fed. Cir. 2009)).
Utilizing the hypothetical negotiation approach, Peter Smith
offers four separate reasonable royalty calculations: (1)
“use Lane Guidance and Turn Preview as a royalty value
proxy;” (2) “use the Steckel part-worths to
calculate potential royalties directly;” (3)
“develop a Total Available Margin estimate;” and
(4) “use the Mitac Settlement.” ECF No. 268,
Garmin Index of Exhibits, Exhibit M, 9/28/12 Expert Report of
Peter Smith 30-31 (“the Smith Report”). In this
motion, Garmin focuses solely on method two, the calculation
that relies on the use of Dr. Steckel's Economic Values,
which Mr. Smith describes as follows:
In this case, I took the partworths developed by Dr. Steckel,
and applied a discount factor to account for the relative
negotiating strength of Visteon, as shown in my development
of the Georgia-Pacific (Exhibits 10 and 11) factors
in this case. That discount, either 0.4000 (Exhibit 10) or
0.568 (Exhibit 11), is based upon the ratio of 2.00/5.00
(Exhibit 10) or 2.84/5.00 (Exhibit 11). It is applied to an
adjusted royalty number developed by assuming a target margin
to be received by Garmin for the incorporated features that
will yield a total margin equal to the actual 63.89% gross
margin received during the infringement period on sales of
automotive PND products. The estimated royalty for each
patent, and the estimated total royalty ($4.94, using the
results of Exhibit 10) to be paid for the four
patents-in-suit is shown in Exhibit 1.
Smith Report 30. (Alteration/correction added; see
also Steckel Dep. 225:3-8).
Smith states that he has read Dr. Steckel's conjoint
study and that it “successfully identified significant
economic value in the patents-in-suit, and can be relied upon
as a partial basis for developing a reasonable royalty
estimate for use in calculating economic damages for use by
the court in the event that it is found Defendant Garmin has
infringed the patents in suit with its PND products.”
ECF No. 268, Exhibit D, Nov. 13, 2012 Deposition of Peter
Smith 54:21-56:8. Mr. Smith confirmed that Dr. Steckel
calculated the consumer value of the accused features only
and did not undertake to determine the value of all features
of the accused devices, which includes a multitude of
features that were not subject to challenge in this case:
Q: Did Dr. Steckel calculate the total consumer value of all
features of personal navigation devices?
A: I believe he only talked about the patented features in
his calculations; and again, in paragraph 69, that is
Q: Okay. So is that correct then, Dr. Steckel did not
calculate the total consumer value of all features, including
non patented features of personal navigation devices?
A: To my knowledge, he didn't and certainly his report
doesn't say that.
Q: [B]ut Dr. Steckel did not determine a parts-worths (sic)
value for all of the non-patented features and
functionalities in the PNDS?
A: That was not the objective of his study, so I'm fairly
certain he didn't. Smith ...