1038-00001 · Cascade Retail Holdings · Pleadings & Court Filings

daubert-motion-keenan.docx DOCUMENT Antitrust

30 / 35 ← → ✕

UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE

DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN

INTRODUCTION

Defendant Cascade Retail Holdings Inc. ("Cascade") respectfully moves this Court, pursuant to Federal Rules of Evidence 702 and 703, and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), and its progeny, to exclude or, in the alternative, to limit the expert testimony of Dr. Wallace Keenan of Thornfield Research Group, the economic expert retained by Plaintiff Federal Trade Commission ("FTC" or "Commission") in connection with this matter.

Dr. Keenan has been designated to testify at the September 8, 2025 evidentiary hearing regarding three principal topics: (1) the definition of the relevant product market; (2) market concentration as measured by the Herfindahl-Hirschman Index ("HHI"); and (3) the predicted competitive effects of Cascade's proposed $4.7 billion acquisition of Pinnacle Market Group Inc. ("Pinnacle"). His opinions are central to the FTC's motion for a preliminary injunction under Section 13(b) of the FTC Act, 15 U.S.C. § 53(b), which seeks to block the transaction pending administrative adjudication before the Commission.

Dr. Keenan's testimony should be excluded in its entirety — or, at a minimum, substantially limited — because his opinions fail to satisfy the reliability requirements of Rule 702 and Daubert in three critical respects:

First, Dr. Keenan's market definition analysis employs a flawed hypothetical monopolist test (commonly referred to as the "SSNIP" test) that systematically ignores supply-side substitution and cross-platform competition. By artificially confining the relevant market to "online marketplace platform services for third-party sellers in the United States," Dr. Keenan excludes the vast majority of competitive alternatives that discipline take rates in the broader e-commerce ecosystem — including direct-to-consumer platforms, social commerce channels, and hybrid retail models. This methodological error is not a mere difference of professional opinion; it is a fundamental analytical flaw that renders his market definition unreliable under any reasonable application of accepted economic principles.

Second, Dr. Keenan's concentration analysis, including his HHI calculations reflecting a pre-merger HHI of 2,361, a post-merger HHI of 2,856, and a delta of 495, is the direct and inescapable product of the improperly narrow market he has defined. Because those concentration figures depend entirely on market shares measured within a gerrymandered market, they are no more reliable than the market definition on which they rest. A properly defined market yields concentration figures that fall far below the thresholds at which the Horizontal Merger Guidelines identify competitive concern.

Third, Dr. Keenan's competitive effects methodology rests on seller-side diversion ratios derived from unreliable survey data and untested assumptions. The survey upon which he relies was designed after this litigation commenced, was framed in a manner that artificially constrained seller choices to marketplace platforms, and has not been validated for representativeness across the approximately 227,000 combined active sellers on the CascadeMarket and Pinnacle platforms. Moreover, Dr. Keenan's upward pricing pressure ("UPP") model fails entirely to account for the pervasive multi-homing behavior of online sellers and ignores $340 million in documented, merger-specific efficiencies.

These are not peripheral quibbles that can be addressed on cross-examination. Dr. Keenan's flawed market definition is the foundation upon which his concentration and competitive effects opinions are constructed, and the unreliability of that foundation infects every conclusion in his expert report. Permitting the FTC to present Dr. Keenan's opinions to this Court at the preliminary injunction hearing — opinions that fail to meet the threshold requirements for admissibility under Rule 702 — would undermine the gatekeeping function that Daubert and its progeny impose.

The evidentiary hearing on the FTC's preliminary injunction motion is set for September 8, 2025. Timely resolution of this motion is necessary to ensure that the hearing proceeds on the basis of reliable evidence and that the parties can prepare accordingly.

I. BACKGROUND

A. The Transaction

On March 14, 2025, Cascade entered into an Agreement and Plan of Merger (the "Merger Agreement") to acquire 100% of the outstanding common stock of Pinnacle for $26.85 per share in cash, representing aggregate consideration of approximately $4.7 billion. The transaction is structured as a reverse triangular merger under Delaware General Corporation Law, pursuant to which a wholly owned acquisition subsidiary of Cascade will merge with and into Pinnacle, with Pinnacle surviving the merger as a wholly owned subsidiary of Cascade.

Cascade is a Delaware corporation headquartered in Seattle, Washington. It operates a diversified omnichannel retail business that includes approximately 340 brick-and-mortar stores located across 38 states, a direct-to-consumer e-commerce platform operating under the brand name CascadeDirect.com, and a third-party seller marketplace known as CascadeMarket, which Cascade launched in June 2019 as part of a broader strategic initiative to expand its digital retail footprint. In fiscal year 2024, CascadeMarket generated approximately $3.1 billion in gross merchandise value ("GMV") from approximately 42,000 active third-party sellers, yielding $527 million in marketplace services revenue at an approximate take rate of 17%.

Pinnacle is a Delaware corporation headquartered in Austin, Texas. Unlike Cascade, Pinnacle operates exclusively as an online marketplace platform, hosting approximately 185,000 active sellers who processed approximately $11.4 billion in GMV in fiscal year 2024. Pinnacle generated approximately $1.82 billion in net revenue during that period, reflecting an approximate take rate of 16%. Pinnacle does not operate any brick-and-mortar retail locations and does not maintain a material direct-to-consumer e-commerce channel separate from its marketplace.

The financing for the proposed acquisition consists of approximately $1.9 billion in available cash on hand from Cascade's balance sheet and a $2.8 billion senior secured term loan B facility from Whitmore National Bank, N.A., as administrative agent and lead arranger. The commitment letter for the term loan facility was executed on March 12, 2025, two days before the execution of the Merger Agreement. The parties filed their premerger notification under the Hart-Scott-Rodino Antitrust Improvements Act of 1976, 15 U.S.C. § 18a, on April 7, 2025, and the Commission issued a Second Request for additional information and documentary material on May 9, 2025.

B. The FTC's Challenge

On June 20, 2025, the Commission — by a vote of 3-2 — authorized the filing of an administrative complaint (FTC Docket No. 9438) challenging the proposed transaction under Section 7 of the Clayton Act, 15 U.S.C. § 18. Three days later, on June 23, 2025, the FTC filed its complaint in this Court seeking a preliminary injunction under Section 13(b) of the FTC Act, 15 U.S.C. § 53(b), to enjoin the consummation of the merger pending the completion of administrative proceedings before the Commission.

The FTC alleges that the relevant product market in which to assess the competitive effects of the merger is "online marketplace platform services for third-party sellers in the United States." Within that proposed market, the FTC alleges the following shares, measured by GMV flowing through each platform's third-party marketplace:

Based on these market shares, the FTC alleges that the proposed acquisition would increase market concentration from a pre-merger HHI of 2,361 to a post-merger HHI of 2,856, representing an increase (delta) of 495. The Commission contends that these concentration levels, coupled with Dr. Keenan's competitive effects analysis, establish a presumption that the merger is likely to substantially lessen competition in the alleged relevant market.

C. Dr. Keenan's Engagement and Opinions

The FTC retained Dr. Wallace Keenan of Thornfield Research Group, located at 1515 Wilson Boulevard, Suite 300, Arlington, Virginia 22209, as its testifying economic expert in this matter. Dr. Keenan submitted his expert report on August 18, 2025, in accordance with the scheduling order entered by this Court. In his report, Dr. Keenan offers opinions on three principal topics: (1) the definition of the relevant product market; (2) market concentration as measured by HHI within the proposed market; and (3) predicted competitive effects of the merger, including an analysis of upward pricing pressure on seller take rates based on seller-side diversion ratios.

Dr. Keenan's opinions rest heavily on a hypothetical monopolist test — the SSNIP test — applied exclusively to third-party marketplace platform services. His competitive effects analysis is derived from seller-side diversion ratios that purport to measure the competitive closeness of CascadeMarket and Pinnacle as substitutes for third-party sellers. As detailed below, each of these analytical pillars suffers from fundamental methodological deficiencies that render Dr. Keenan's opinions unreliable and inadmissible under Rule 702 and Daubert.

II. LEGAL STANDARD

A. Federal Rule of Evidence 702

Federal Rule of Evidence 702, as amended effective December 1, 2023, provides:

A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that:

(a) 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;

(b) the testimony is based on sufficient facts or data;

(d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.

Fed. R. Evid. 702. The 2023 amendment clarified the preponderance-of-the-evidence standard that governs the admissibility determination, emphasizing that "the proponent" — here, the FTC — "bears the burden of establishing" each element of admissibility. See Fed. R. Evid. 702 advisory committee's note to 2023 amendment ("The amendment clarifies that the preponderance standard applies to the Rule 702(d) determination as well, and that an opinion is not reliable simply because the expert says it is."). This amendment was specifically intended to address the concern that some courts were too readily admitting expert testimony by conflating the admissibility inquiry with the assessment of the weight or credibility of the testimony.

B. Daubert and Its Progeny

In Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589 (1993), the Supreme Court held that Federal Rule of Evidence 702 imposes upon the trial court a "gatekeeping" obligation to ensure that expert testimony is both relevant and reliable before it may be admitted. The Court identified several non-exclusive factors that may bear on the reliability of expert testimony, including: (1) whether the theory or technique can be and has been tested; (2) whether it has been subjected to peer review and publication; (3) the known or potential rate of error; and (4) general acceptance within the relevant scientific or professional community. Id. at 593-94. These factors are "meant to be helpful, not definitive," and courts are afforded "considerable leeway in deciding in a particular case how to go about determining whether particular expert testimony is reliable." Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999).

In General Electric Co. v. Joiner, 522 U.S. 136, 146 (1997), the Court emphasized that the gatekeeping function requires courts to examine whether there is "too great an analytical gap between the data and the opinion proffered." An expert opinion that is connected to existing data or methodology "only by the ipse dixit of the expert" is insufficient to satisfy the reliability standard. Id.

In Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), the Court extended the Daubert gatekeeping obligation to all forms of expert testimony, including non-scientific testimony based on "technical" or "other specialized" knowledge. Id. at 141. This holding is of particular relevance here, as Dr. Keenan's testimony rests on economic methodology rather than traditional scientific experimentation, but is no less subject to the Court's gatekeeping scrutiny.

C. Application to Economic Expert Testimony in Antitrust Cases

Courts presiding over merger challenges routinely exercise their gatekeeping function under Rule 702 and Daubert to exclude or limit the testimony of economic experts whose methodologies are unreliable or whose opinions are not grounded in a sound application of economic principles to the facts of the case. See, e.g., FTC v. Sysco Corp., 113 F. Supp. 3d 1, 25-26 (D.D.C. 2015) (evaluating reliability of economic expert's market definition analysis); United States v. H&R Pinnacle Holdings, 833 F. Supp. 2d 36, 72-73 (D.D.C. 2011) (same); In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 323 (3d Cir. 2008) (holding that a trial court must "resolve all Daubert issues before certifying a class," including the reliability of economic expert opinions on market definition).

Particular scrutiny is warranted where, as here, the expert's market definition is the keystone of the entire competitive effects analysis. Errors or unreliable assumptions embedded in the market definition propagate through the concentration analysis and into the competitive effects conclusions, amplifying the impact of any initial methodological flaw. See ABA Section of Antitrust Law, Mergers and Acquisitions 167 (4th ed. 2016) ("Market definition is often the central battleground in merger litigation because it determines market shares, concentration levels, and the identity of market participants."). As the Ninth Circuit has instructed, expert testimony that "offers no reliable basis on which to analyze" the economic question at issue is properly excluded. Daubert v. Merrell Dow Pharm., Inc., 43 F.3d 1311, 1319 (9th Cir. 1995) (on remand); see also Mukhtar v. Cal. State Univ., Hayward, 299 F.3d 1053, 1066 (9th Cir. 2002) ("[A]n expert who supplies nothing but a bottom line supplies nothing of value to the judicial process.").

III. ARGUMENT

A. Dr. Keenan's Market Definition Analysis Is Unreliable Because It Employs a Flawed SSNIP Test

Dr. Keenan defines the relevant product market as "online marketplace platform services for third-party sellers in the United States." This narrow market definition is the product of a hypothetical monopolist test that is fundamentally flawed in its design and execution. Specifically, Dr. Keenan's SSNIP analysis systematically ignores supply-side substitution and cross-platform competition, artificially constraining the relevant market and ensuring a result that supports the FTC's enforcement theory. These are not minor methodological disagreements; they are foundational errors that render Dr. Keenan's market definition unreliable and inadmissible.

1. The Hypothetical Monopolist / SSNIP Framework Requires Consideration of All Competitive Alternatives. The hypothetical monopolist test, also known as the SSNIP (Small but Significant and Non-transitory Increase in Price) test, is the standard analytical framework for defining relevant antitrust markets. Under this test, the analyst begins with a candidate market consisting of the products (or services) of the merging firms and asks whether a hypothetical monopolist of all products in the candidate market could profitably impose a small but significant and non-transitory increase in price — typically 5% to 10% — over prevailing levels. If a sufficient number of customers (or, on the sell side, suppliers) would switch to alternative products outside the candidate market in response to such a price increase, making it unprofitable, the candidate market must be broadened to include the next-best substitute. The process is repeated iteratively until the hypothetical monopolist could profitably sustain the price increase. See U.S. Dep't of Justice & FTC, Horizontal Merger Guidelines § 4.1.1 (2010); FTC & U.S. Dep't of Justice, Merger Guidelines § 4.3 (2023).

Critically, a properly conducted SSNIP test must account for both demand-side and supply-side substitution. Demand-side substitution refers to the willingness of customers (here, sellers) to switch to competing products or services in response to a price increase. Supply-side substitution refers to the ability of firms not currently participating in the candidate market to rapidly reposition their offerings to enter the market in response to a price increase, thereby constraining the hypothetical monopolist's ability to raise prices profitably. Both the 2010 and 2023 Horizontal Merger Guidelines expressly recognize that supply-side substitution can discipline prices just as effectively as demand-side switching and should be incorporated into the market definition analysis. See 2010 Guidelines § 5.1 ("The Agencies consider the competitive significance of possible supply responses in defining markets and measuring market shares."); 2023 Guidelines § 4.3.C (recognizing that "rapid supply-side entry" may constrain market power).

The failure to account for either dimension of substitution in a SSNIP test produces a candidate market that is artificially narrow and therefore unreliable. An artificially narrow market, by definition, overstates the market power of the merging firms and exaggerates the competitive significance of the proposed transaction.

2. Dr. Keenan's SSNIP Test Ignores Supply-Side Substitution. Dr. Keenan defined the candidate market as "online marketplace platform services for third-party sellers" and conducted his hypothetical monopolist test by examining only whether third-party sellers would switch to other dedicated marketplace platforms in response to a take-rate increase. Keenan Report ¶¶ 42-58. His analysis entirely omitted the supply side — that is, the demonstrated ability and economic incentive of firms operating outside the candidate market to rapidly offer marketplace-type services to third-party sellers in response to a supra-competitive increase in take rates.

The evidence of supply-side substitution is substantial and readily available. At least three categories of firms are positioned to offer, and in several cases have already begun offering, marketplace-type services to third-party sellers:

Direct-to-consumer e-commerce platforms. Cascade itself operates CascadeDirect.com, a direct-to-consumer e-commerce platform that generated approximately $40.6 billion in sales in FY2024. The very existence of CascadeMarket — launched in 2019 by repurposing Cascade's existing e-commerce infrastructure — demonstrates that a direct-to-consumer platform can be repositioned to serve third-party sellers at relatively low marginal cost and within a short time frame. Other major e-commerce retailers, including at least three firms with revenues exceeding $5 billion, operate or are developing similar third-party seller capabilities. Dr. Keenan's SSNIP test entirely ignores these competitive alternatives.

Social commerce platforms. The emergence of social commerce represents a rapidly growing competitive alternative to dedicated marketplace platforms. Brevity Social Commerce, launched in 2023, enables third-party sellers to list and sell products directly through integrated social media storefronts. Social commerce platforms generated an estimated $67 billion in U.S. GMV in 2024, a figure projected to grow at a compound annual rate of approximately 25% through 2028. These platforms compete directly with marketplace platforms for third-party seller volume, yet Dr. Keenan excluded them from his analysis without providing any economic justification.

Hybrid retail models. Trove Marketplace, launched in 2024, represents a new category of hybrid retail platform that combines elements of traditional e-commerce, social commerce, and marketplace functionality. Trove and similar platforms allow third-party sellers to access consumers through multiple channels within a single integrated environment. The entry of Trove and comparable platforms demonstrates that the barriers to supply-side entry into marketplace services for third-party sellers are low and that repositioning by adjacent competitors is both feasible and occurring in the marketplace today.

By ignoring supply-side substitution from these and other competitive alternatives, Dr. Keenan artificially constrained the set of products and services included in his candidate market and ensured that his SSNIP test would yield a narrow market limited to dedicated marketplace platforms. This is a results-driven methodological choice, not an analytically sound application of the hypothetical monopolist framework. It renders Dr. Keenan's market definition unreliable under Rule 702 and Daubert.

3. Dr. Keenan's SSNIP Test Ignores Cross-Platform Competition. Dr. Keenan's SSNIP test is further undermined by his failure to account for the pervasive cross-platform competition that characterizes the e-commerce sector. Online sellers routinely engage in "multi-homing" — that is, they sell their products across multiple platforms simultaneously, distributing their sales volume among marketplace platforms, direct-to-consumer storefronts, social media commerce channels, and hybrid platforms.

The evidence of multi-homing is extensive and directly relevant to the SSNIP analysis. Internal Cascade documents produced in discovery, as well as internal Pinnacle documents, demonstrate that a substantial majority of active sellers on both CascadeMarket and Pinnacle simultaneously maintain direct-to-consumer e-commerce storefronts, sell through one or more social media commerce channels, or utilize hybrid platforms. Publicly available industry surveys corroborate this finding. For example, the 2024 E-Commerce Seller Survey published by the National Retail Federation found that 73% of sellers active on at least one online marketplace also sell through at least one non-marketplace channel. Similarly, Pinnacle's own internal data, as reflected in its FY2024 10-K filing, notes that "a significant portion of our active sellers maintain selling presences on multiple platforms, including direct-to-consumer e-commerce sites and social commerce channels."

A proper SSNIP analysis must account for this cross-platform competition because multi-homing means that a take-rate increase on one platform causes marginal sellers to shift volume — not necessarily their entire business — to competing channels. The relevant constraint on a hypothetical monopolist's ability to raise take rates profitably is not limited to dedicated marketplace platforms; it includes every alternative channel through which a seller can reach consumers and generate sales. A seller who maintains a storefront on Pinnacle, a direct-to-consumer website, and a Brevity Social Commerce shop does not need to abandon Pinnacle entirely in response to a take-rate increase. Instead, the seller can shift marginal advertising spend, inventory allocation, and promotional activity toward competing channels — a form of competitive discipline that is invisible to Dr. Keenan's marketplace-only SSNIP test.

Dr. Keenan treated marketplace platforms as though they exist in a competitive vacuum, isolated from the broader $757 billion U.S. e-commerce ecosystem (2024 estimate, U.S. Census Bureau). This assumption is contradicted by overwhelming evidence of cross-platform competition and multi-homing. Cascade's expert, Dr. Priya Ramanathan of Brindoval Analytics LLC, demonstrates in her analysis that when cross-platform substitution is properly accounted for, the relevant market encompasses all e-commerce retail platforms competing for seller volume. In that properly defined market, the combined entity's share is approximately 7.3% of total U.S. e-commerce GMV — a share level at which no plausible theory of competitive harm can be sustained.

4. The Analytical Gap Between Dr. Keenan's Data and His Market Definition Conclusion. Even accepting Dr. Keenan's own analytical framework arguendo, the data he assembled do not support the narrow market definition he reached. This analytical gap is precisely the kind of disconnect between data and opinion that the Supreme Court identified in Joiner as grounds for exclusion.

Dr. Keenan's own survey data reveal that 34% of sellers surveyed indicated that they would consider shifting sales volume to non-marketplace channels — including direct-to-consumer websites and social commerce platforms — in response to a take-rate increase of the magnitude specified in his hypothetical monopolist test. Keenan Report, Appendix C, Table 5. Under standard SSNIP methodology, this level of switching would ordinarily compel the analyst to expand the candidate market to include those non-marketplace channels, because the diversion of even a fraction of this volume could render the hypothetical monopolist's price increase unprofitable.

Rather than following the logic of his own data, Dr. Keenan disregarded this evidence by applying an ad hoc "critical loss" threshold that is inconsistent with his own stated methodology. Specifically, Dr. Keenan asserted in his report that sellers who indicated a willingness to shift volume to non-marketplace channels should be disregarded because, in his judgment, those sellers were "unlikely to complete a full transition" away from marketplace platforms. Keenan Report ¶ 54. But the SSNIP test does not require a "full transition"; it requires only that sufficient volume shift to make the price increase unprofitable. By imposing a full-transition requirement that has no basis in accepted economic methodology, Dr. Keenan effectively moved the goalposts to ensure that his SSNIP test produced the narrow market definition the FTC's theory demands.

This is the paradigmatic "analytical gap" between data and conclusion that the Joiner Court warned against. Dr. Keenan's data point in one direction — toward a broader market — but his conclusion points in another. The gap is bridged not by rigorous economic analysis but by the ipse dixit of the expert. Under Daubert and Rule 702, this gap warrants exclusion.

B. Dr. Keenan's HHI Concentration Analysis Is Unreliable Because It Is Derived From the Flawed Market Definition

1. The HHI Calculations Are Only as Reliable as the Underlying Market. Dr. Keenan calculates a pre-merger HHI of 2,361, a post-merger HHI of 2,856, and an HHI increase (delta) of 495. The HHI delta is derived mechanically from the market shares of CascadeMarket and Pinnacle within the proposed market: 2 × 8.2 × 30.1 = 493.64, which the FTC rounds to 495 in its Complaint at ¶ 47.

These figures are derived exclusively from market shares measured within the FTC's proposed narrow market of "online marketplace platform services for third-party sellers in the United States." The share figures — Cascade 8.2%, Pinnacle 30.1%, Vertex Commerce 35.7%, Nomad Shops 14.6%, and all others 11.4% — are meaningful only within the confines of that particular market definition. If the market definition is unreliable, as Cascade has demonstrated above, then the HHI calculations constructed upon it are necessarily unreliable as well.

The principle is straightforward: "garbage in, garbage out." A mathematically precise HHI computation does not cure a fundamentally flawed market definition input. The HHI is merely a summary statistic that describes the distribution of market shares within a defined market. If the market is defined too narrowly — as it is here — then the shares allocated to each firm will be inflated relative to shares that would prevail in a properly defined market, and the resulting HHI will overstate the degree of concentration. The precision of the arithmetic is irrelevant if the inputs to the arithmetic are unreliable. See FTC v. Sanford Health, 926 F.3d 959, 965 (8th Cir. 2019) (recognizing that market definition determines the universe of competitors whose shares are measured).

2. A Properly Defined Market Yields Dramatically Different Concentration Results. The magnitude of the distortion introduced by Dr. Keenan's narrow market definition is readily apparent when concentration is measured in a properly defined market. In a market defined as all U.S. e-commerce retail — which, as set forth above, encompasses the full range of competitive alternatives available to third-party sellers — total GMV in 2024 was approximately $757 billion (U.S. Census Bureau estimate).

Within that broader market, the combined entity's marketplace GMV of approximately $14.5 billion ($3.1 billion from CascadeMarket plus $11.4 billion from Pinnacle) represents approximately 1.9% of the market. Even including Cascade's total e-commerce GMV — approximately $43.7 billion, including both direct-to-consumer sales through CascadeDirect.com and third-party marketplace sales through CascadeMarket — the combined entity's share is approximately 7.3% ($55.1 billion divided by $757 billion).

At these share levels, the post-merger HHI would fall far below the thresholds at which any version of the Horizontal Merger Guidelines identifies competitive concern. Under both the 2010 and 2023 Guidelines, mergers that result in post-merger HHI levels below 1,500 are generally considered unlikely to raise competitive concerns, and mergers resulting in HHI levels between 1,500 and 2,500 are considered potentially concerning only if the delta exceeds certain thresholds. A combined share of 7.3% in a market with numerous large competitors would not approach these thresholds.

Dr. Keenan never analyzed concentration in the broader market. He did not compute HHI levels for any market definition other than the narrow one the FTC's theory requires. This omission further demonstrates the outcome-determinative nature of his narrow market definition: Dr. Keenan's entire analytical framework is designed to produce concentration figures that support a presumption of illegality, rather than to conduct an objective assessment of the competitive significance of the proposed transaction.

3. Dr. Keenan's Market Share Measurement Methodology Contains Additional Flaws. Even within the narrow market Dr. Keenan has defined, his methodology for measuring market shares contains additional errors that compound the market definition problem and render his HHI calculations even less reliable.

First, Dr. Keenan measured shares by GMV flowing through the seller side of each marketplace platform, but failed to account for the two-sided nature of the platform. Online marketplace platforms are classic two-sided transaction platforms: they serve both sellers (who list products and pay take rates) and consumers (who purchase products). The Supreme Court's analysis in Ohio v. Halcyon Corporation, 138 S. Ct. 2274 (2018), instructs that both sides of a two-sided transaction platform must be considered in assessing competitive effects. While Amex addressed the substantive antitrust analysis rather than Rule 702 admissibility, its recognition that both sides of a platform matter is directly relevant to the reliability of Dr. Keenan's one-sided market share methodology. A market share analysis that accounts for seller-side GMV but ignores consumer-side transaction volume and competitive dynamics produces a distorted picture of competitive significance.

Second, Dr. Keenan excluded from his share calculations all transaction volume that flows through non-marketplace e-commerce channels operated by the same firms whose marketplace shares he measured. Most notably, he measured Cascade's share using only the $3.1 billion in GMV from CascadeMarket, while excluding the approximately $40.6 billion in direct-to-consumer sales through CascadeDirect.com. This exclusion is analytically problematic because CascadeDirect.com competes directly with marketplace platforms for seller volume (through vendor relationships and wholesale purchasing) and for consumer purchasing. By measuring Cascade's share as only 8.2% rather than a figure reflecting its total e-commerce volume, Dr. Keenan's methodology paradoxically understates Cascade's competitive significance in some respects while simultaneously overstating the merger's impact on concentration by confining the analysis to a narrow market in which the combination appears more consequential.

These measurement errors compound the market definition error described in Section III.A and render the HHI figures presented in Dr. Keenan's report unreliable as a foundation for expert testimony.

C. Dr. Keenan's Competitive Effects Analysis Relies on Unreliable Seller-Side Diversion Ratios

1. Description of Dr. Keenan's Diversion Ratio Methodology. Beyond his market definition and concentration opinions, Dr. Keenan offers opinions regarding the predicted competitive effects of the merger. The centerpiece of his competitive effects analysis is a set of seller-side diversion ratios purporting to demonstrate that CascadeMarket and Pinnacle are "close competitors" — that is, that a significant share of the sellers who would leave one platform in response to a take-rate increase would switch to the other platform. Keenan Report ¶¶ 72-91.

Dr. Keenan used these diversion ratios as inputs to an upward pricing pressure ("UPP") model. The UPP framework estimates the incentive of the merged firm to raise price (here, take rates) by comparing the value of sales recaptured through the merger partner (measured by the diversion ratio multiplied by the margin earned on diverted sales) against any countervailing downward pressure from merger-specific efficiencies. Dr. Keenan's UPP analysis purports to show that the merged entity would have an incentive to raise take rates for third-party sellers on both platforms, resulting in competitive harm. Keenan Report ¶¶ 92-108.

The reliability of Dr. Keenan's competitive effects analysis depends entirely on the reliability of his diversion ratio estimates, which in turn depend on the reliability of the survey data from which they are derived. As set forth below, both the survey data and the diversion ratio methodology are fatally flawed.

2. The Underlying Survey Data Are Unreliable. Dr. Keenan's seller-side diversion ratios are based on a survey of third-party sellers conducted by Thornfield Research Group. This survey suffers from fundamental design flaws that render it unreliable as a foundation for expert testimony under Rule 702 and Daubert.

Litigation-driven framing. The survey was designed and administered after the FTC filed its administrative complaint on June 20, 2025. Keenan Report, Appendix D (survey methodology) (survey fielded July 7-21, 2025). While the timing of a survey does not automatically disqualify it, the post-litigation design of the survey raises serious concerns that it was framed to generate results supportive of the FTC's enforcement theory rather than to objectively measure seller switching behavior. Courts have recognized that surveys designed in the context of pending litigation warrant heightened scrutiny. See Shari Scharlin, Survey Evidence in Antitrust Proceedings, 23 Antitrust L.J. 355 (2019).

Artificially constrained choice set. The survey presented sellers with a hypothetical question about their switching behavior in response to a take-rate increase, but the hypothetical was framed in a manner that limited seller response options to other marketplace platforms. Specifically, the question asked: "If [platform name] increased its take rate by [X]%, to which online marketplace platform would you most likely shift your sales volume?" Keenan Report, Appendix D, Survey Instrument, Q12. By embedding the phrase "online marketplace platform" in the question itself, the survey effectively excluded direct-to-consumer and social commerce alternatives from the choice set. Sellers who would have indicated a preference for shifting volume to their own website, to a social commerce channel, or to a hybrid platform were not given the opportunity to express that preference. The resulting diversion ratios are therefore biased upward toward marketplace-to-marketplace diversion, precisely the outcome the FTC's theory requires.

Inadequate validation for representativeness. The combined active seller base on CascadeMarket and Pinnacle totals approximately 227,000 sellers (42,000 on CascadeMarket and 185,000 on Pinnacle). Dr. Keenan does not disclose the total number of survey responses or the response rate in the body of his report, noting only in Appendix D that the survey achieved "a sufficient number of responses for statistical analysis." Keenan Report, Appendix D, at 3. Nor does Dr. Keenan demonstrate that the survey sample is representative of the seller population along critical dimensions, including seller size (by GMV), product category, tenure on the platform, and degree of multi-homing. Without adequate validation for representativeness, the survey cannot be relied upon to generate diversion ratio estimates that are generalizable to the seller population as a whole.

An expert opinion constructed on the foundation of unreliable survey data is itself unreliable under Rule 702 and Daubert. Where, as here, the survey instrument contains design flaws that bias the results in a predictable direction, the survey cannot serve as a reliable basis for the expert's conclusions.

3. The Diversion Ratio Model Ignores Multi-Homing. Even if Dr. Keenan's survey data were reliable — and they are not — his diversion ratio framework would still be analytically unsound because it fails to account for the multi-homing behavior of online sellers.

As discussed in Section III.A.3 above, a substantial majority of sellers on both CascadeMarket and Pinnacle sell simultaneously on multiple platforms, including marketplace platforms, direct-to-consumer websites, and social commerce channels. Multi-homing is not an incidental feature of the e-commerce landscape; it is a defining characteristic that fundamentally shapes seller behavior and the competitive dynamics of the industry.

Multi-homing undermines the analytical coherence of the diversion ratio framework in several ways. Most fundamentally, if a seller already sells on both CascadeMarket and Pinnacle — as many sellers do — the concept of "diversion" from one platform to the other is analytically incoherent. A seller who is present on both platforms cannot be "diverted" to a platform where it already operates. What the seller can do, in response to a take-rate increase on one platform, is shift marginal volume toward the other platform (or toward other channels), but this marginal volume shift is conceptually and quantitatively distinct from the wholesale diversion of a seller's business that Dr. Keenan's model assumes.

Dr. Keenan's diversion ratio model fails to distinguish between diversion of marginal volume and diversion of an entire seller relationship. By treating every seller as though it either sells on CascadeMarket or Pinnacle but not both, the model overstates the degree to which a take-rate increase on one platform would benefit the other, leading to inflated diversion ratio estimates that overstate the merged firm's incentive to raise take rates.

Cascade's rebuttal expert, Dr. Priya Ramanathan of Brindoval Analytics LLC, presents an alternative econometric model in her expert report that explicitly accounts for seller multi-homing. Dr. Ramanathan's analysis demonstrates that when multi-homing rates are properly incorporated into the competitive effects framework, the predicted diversion from CascadeMarket to Pinnacle (and vice versa) drops substantially, rendering the FTC's unilateral effects theory implausible. Dr. Ramanathan's rebuttal report, due August 29, 2025 under the Court's scheduling order, will be submitted in advance of the evidentiary hearing and will provide the Court with a methodologically sound alternative to Dr. Keenan's flawed analysis.

4. The UPP Analysis Fails to Account for Efficiencies. Even setting aside the deficiencies in Dr. Keenan's diversion ratios, his upward pricing pressure analysis is independently unreliable because it fails to account for cognizable, merger-specific efficiencies.

A properly specified UPP model incorporates merger-specific efficiencies as a downward offset against the upward pricing pressure generated by the internalization of competitive externalities between the merging firms. See Joseph Farrell & Carl Shapiro, Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition, 10 B.E. J. Theoretical Econ. 1, 9-10 (2010) (describing the role of efficiencies as a "default efficiency" offset in the UPP framework). The Horizontal Merger Guidelines similarly recognize that "cognizable efficiencies" — those that are merger-specific, verifiable, and not derived from anticompetitive reductions in output or service — should be credited in the competitive effects analysis. 2010 Guidelines § 10; 2023 Guidelines § 3.4.

Cascade has documented $340 million in annual merger-specific efficiencies projected to be achieved by Year 3 following consummation of the transaction. These efficiencies include:

• Logistics and fulfillment cost savings: $195 million. Integration of Pinnacle's third-party sellers into Cascade's existing 22-warehouse fulfillment network, which currently serves CascadeMarket sellers as well as Cascade's direct-to-consumer business. Cascade's fulfillment infrastructure has significant excess capacity that can absorb Pinnacle's seller volume at marginal cost. The resulting efficiencies include reduced per-unit shipping costs, consolidation of inbound logistics, and elimination of redundant fulfillment center leases currently maintained by Pinnacle's third-party logistics partners.

• Technology and platform synergies: $145 million. Consolidation of duplicative technology platforms, including seller management tools, search and discovery algorithms, payment processing systems, and customer service infrastructure. Cascade's investment of over $800 million in its e-commerce technology stack since 2019 provides a scalable foundation that can be extended to Pinnacle's seller base without proportional incremental cost.

These efficiencies are merger-specific (they cannot be achieved absent the transaction), verifiable (they are supported by detailed integration planning documents and third-party validation by Cascade's financial advisor), and would be passed through to sellers in the form of lower costs and improved services.

Dr. Keenan's UPP analysis either ignores these efficiencies entirely or dismisses them as "non-cognizable" without conducting any independent verification analysis. Keenan Report ¶¶ 105-108. Dr. Keenan does not examine Cascade's integration planning documents, does not interview Cascade's operations team, and does not apply any recognized methodology for evaluating whether claimed efficiencies are merger-specific and verifiable. He simply asserts, in a conclusory fashion, that "the claimed efficiencies are speculative and unverified." Keenan Report ¶ 107. This is not economic analysis; it is advocacy. A UPP model that ignores one of its essential inputs — the efficiency offset — is incomplete and unreliable, and the resulting competitive effects opinions should be excluded.

D. The Flaws in Dr. Keenan's Analysis Are Not Matters of Weight But of Admissibility

Cascade anticipates that the FTC will respond to this motion by arguing that the deficiencies identified above go merely to the "weight, not the admissibility" of Dr. Keenan's testimony, and that they can be adequately addressed through vigorous cross-examination at the evidentiary hearing. This argument should be rejected.

The distinction between weight and admissibility is a meaningful one, but it does not operate to excuse fundamental methodological errors of the kind present in Dr. Keenan's analysis. A disagreement between competing experts about the correct value of an input parameter, or about the relative probative value of different data sources, is typically a matter of weight. But the wholesale omission of an analytically essential category of competitive alternatives (supply-side substitution and cross-platform competition) from a SSNIP test, the use of survey data with a biased choice set, and the failure to account for a defining feature of the competitive landscape (multi-homing) are not differences of degree — they are differences of kind. They reflect not a reasonable exercise of expert judgment but a systematic departure from accepted economic methodology.

Moreover, the errors in Dr. Keenan's analysis are not isolated. They are interconnected and mutually reinforcing. The narrow market definition produces inflated market shares, which produce inflated HHI levels, which create an appearance of high concentration that is then compounded by the inflated diversion ratios derived from the flawed survey. The entire chain of reasoning is built on an unreliable foundation, and removing any one link does not cure the deficiency because the remaining links are themselves infected by the same foundational errors.

The 2023 amendment to Rule 702 was specifically intended to correct the tendency of some courts to "admit expert testimony too readily" by conflating the admissibility determination with the assessment of weight. Fed. R. Evid. 702 advisory committee's note to 2023 amendment. The amendment reinforces that "the proponent must demonstrate to the court that it is more likely than not that the proffered testimony meets the admissibility requirements set out in the rule." Id. Here, the FTC cannot carry that burden, because the methodological flaws in Dr. Keenan's analysis are sufficiently fundamental to render his opinions unreliable under any reasonable application of Rule 702 and Daubert.

Under Daubert and its progeny, it is the Court's responsibility as gatekeeper to exclude expert testimony that fails to meet reliability thresholds. The gatekeeping function exists precisely to prevent unreliable evidence from being presented to the finder of fact, because once such evidence has been received, the damage cannot always be undone through cross-examination. As the Supreme Court recognized, "vigorous cross-examination" and "presentation of contrary evidence" are not adequate substitutes for the Court's independent assessment of reliability. Daubert, 509 U.S. at 596. Rather than leaving it to Cascade to "unring the bell" through cross-examination, the Court should exercise its gatekeeping authority to exclude Dr. Keenan's unreliable testimony.

E. In the Alternative, the Court Should Limit Dr. Keenan's Testimony

Should the Court decline to exclude Dr. Keenan's testimony in its entirety, Cascade respectfully requests, in the alternative, that the Court enter an order limiting the scope of Dr. Keenan's testimony at the September 8, 2025 evidentiary hearing as follows:

First, the Court should preclude Dr. Keenan from testifying to the conclusion that the relevant product market is limited to "online marketplace platform services for third-party sellers in the United States." This conclusion is the product of a SSNIP test that systematically excludes supply-side substitution and cross-platform competition, and it does not meet the reliability threshold required by Rule 702. Dr. Keenan should be permitted to testify regarding economic principles generally, but not to offer the specific market definition opinion that is the product of his flawed methodology.

Second, the Court should preclude Dr. Keenan from offering HHI calculations — specifically, his figures of 2,361 (pre-merger), 2,856 (post-merger), and 495 (delta) — based solely on market shares within the narrow proposed market. In the alternative, the Court should require Dr. Keenan to present alternative HHI calculations based on a properly defined broader market that accounts for all relevant competitive alternatives, so that the Court may assess concentration in its proper competitive context.

Third, the Court should preclude Dr. Keenan from offering competitive effects opinions derived from seller-side diversion ratios generated by the flawed survey methodology described above. Because the diversion ratios are the product of a survey with an artificially constrained choice set and inadequate validation for representativeness, and because the diversion ratio model fails to account for multi-homing, the competitive effects opinions built upon them do not satisfy the reliability requirements of Rule 702.

These limitations would permit the FTC to present whatever reliable economic testimony Dr. Keenan can offer, while preventing the Court from being presented with unreliable opinions that do not satisfy the threshold requirements for admissibility.

IV. CONCLUSION

For the foregoing reasons, Defendant Cascade Retail Holdings Inc. respectfully requests that this Court grant its motion and enter an order excluding the expert testimony of Dr. Wallace Keenan in its entirety or, in the alternative, limiting his testimony as specified in Section III.E above.

This motion is timely filed in advance of the evidentiary hearing on the FTC's motion for a preliminary injunction, which is set for September 8, 2025. Cascade respectfully submits that resolution of this motion prior to the hearing is necessary to ensure that the hearing proceeds on the basis of reliable evidence and that the parties and the Court can allocate hearing time accordingly.

Cascade respectfully requests oral argument on this motion if the Court deems it appropriate.

Respectfully submitted,

CALDERWOOD & HARKNESS LLP

By: ________

Howard Bellamy D.C. Bar No. 487231 Calderwood & Harkness LLP 1401 K Street NW, Suite 1200 Washington, D.C. 20005 Telephone: (202) 555-4100 Facsimile: (202) 555-4101

Email: hbellamy@calderwoodharkness.com

Aaron Liu Cal. Bar No. 321574 Calderwood & Harkness LLP 555 California Street, Suite 2700 San Francisco, CA 94104 Telephone: (415) 555-6200 Facsimile: (415) 555-6201

Email: aliu@calderwoodharkness.com

Counsel for Defendant Cascade Retail Holdings Inc.

Matter No.: 1038-00001

Dated: August 25, 2025

CERTIFICATE OF SERVICE

I hereby certify that on August 25, 2025, I caused a true and correct copy of Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan, together with all supporting papers, to be served upon the following counsel of record via the Court's CM/ECF electronic filing system, which will send notification of such filing to all registered participants:

Counsel for Plaintiff Federal Trade Commission:

Claire Dunmore, Lead Trial Attorney Samuel Ortega, Senior Trial Counsel Rebecca Yun, Staff Attorney Federal Trade Commission Bureau of Competition, Mergers IV Division 600 Pennsylvania Avenue NW Washington, D.C. 20580

Counsel for Co-Defendant Pinnacle Market Group Inc.:

Diana Whitfield Kenneth Oshiro Whitfield, Greer & Tanaka LLP 100 Congress Avenue, Suite 2200 Austin, TX 78701

Counsel for Proposed Intervenor Independent Online Sellers Alliance (IOSA):

Martina Galloway Galloway & Peck LLP 60 East 42nd Street, Suite 2600 New York, NY 10165

________________________________________

Howard Bellamy Calderwood & Harkness LLP

UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE

[PROPOSED] ORDER GRANTING DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN

This matter comes before the Court on Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan, filed August 25, 2025. The Court has considered the motion and all supporting papers, the opposition filed by Plaintiff Federal Trade Commission, and the reply filed by Defendant Cascade Retail Holdings Inc. Having considered the submissions of the parties, the relevant legal authorities, and the full record in this matter, the Court finds as follows:

IT IS HEREBY ORDERED that Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan is GRANTED [or GRANTED IN PART].

IT IS FURTHER ORDERED that Dr. Wallace Keenan is precluded from offering the following testimony at the September 8, 2025 preliminary injunction evidentiary hearing:

________________________________________

SO ORDERED.

Dated: ________

________________________________________

THE HONORABLE RICHARD A. JOHANNSEN United States District Judge

FEDERAL TRADE COMMISSION,	Case No. 2:25-cv-01087-RAJ
<!-- indent:2 -->Plaintiff,<!-- /indent -->	Before the Honorable Richard A. Johannsen
v.	DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN

CASCADE RETAIL HOLDINGS INC. and PINNACLE MARKET GROUP INC.,

<!-- indent:2 -->Defendants.<!-- /indent -->	Evidentiary Hearing: September 8, 2025
Firm	GMV (FY2024)	Market Share
Vertex Commerce Inc.	$13.5 billion	35.7%
Pinnacle Market Group Inc.	$11.4 billion	30.1%
Nomad Shops LLC	$5.5 billion	14.6%
Cascade Retail Holdings Inc. (CascadeMarket)	$3.1 billion	8.2%
All Others	$4.3 billion	11.4%
Total	$37.8 billion	100.0%
FEDERAL TRADE COMMISSION,	Case No. 2:25-cv-01087-RAJ

Plaintiff,

CASCADE RETAIL HOLDINGS INC. and PINNACLE MARKET GROUP INC.,

Defendants.

UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE
DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN
INTRODUCTION
Defendant Cascade Retail Holdings Inc. ("Cascade") respectfully moves this Court, pursuant to Federal Rules of Evidence 702 and 703, and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), and its progeny, to exclude or, in the alternative, to limit the expert testimony of Dr. Wallace Keenan of Thornfield Research Group, the economic expert retained by Plaintiff Federal Trade Commission ("FTC" or "Commission") in connection with this matter.
Dr. Keenan has been designated to testify at the September 8, 2025 evidentiary hearing regarding three principal topics: (1) the definition of the relevant product market; (2) market concentration as measured by the Herfindahl-Hirschman Index ("HHI"); and (3) the predicted competitive effects of Cascade's proposed $4.7 billion acquisition of Pinnacle Market Group Inc. ("Pinnacle"). His opinions are central to the FTC's motion for a preliminary injunction under Section 13(b) of the FTC Act, 15 U.S.C. § 53(b), which seeks to block the transaction pending administrative adjudication before the Commission.
Dr. Keenan's testimony should be excluded in its entirety — or, at a minimum, substantially limited — because his opinions fail to satisfy the reliability requirements of Rule 702 and Daubert in three critical respects:
First, Dr. Keenan's market definition analysis employs a flawed hypothetical monopolist test (commonly referred to as the "SSNIP" test) that systematically ignores supply-side substitution and cross-platform competition. By artificially confining the relevant market to "online marketplace platform services for third-party sellers in the United States," Dr. Keenan excludes the vast majority of competitive alternatives that discipline take rates in the broader e-commerce ecosystem — including direct-to-consumer platforms, social commerce channels, and hybrid retail models. This methodological error is not a mere difference of professional opinion; it is a fundamental analytical flaw that renders his market definition unreliable under any reasonable application of accepted economic principles.
Second, Dr. Keenan's concentration analysis, including his HHI calculations reflecting a pre-merger HHI of 2,361, a post-merger HHI of 2,856, and a delta of 495, is the direct and inescapable product of the improperly narrow market he has defined. Because those concentration figures depend entirely on market shares measured within a gerrymandered market, they are no more reliable than the market definition on which they rest. A properly defined market yields concentration figures that fall far below the thresholds at which the Horizontal Merger Guidelines identify competitive concern.
Third, Dr. Keenan's competitive effects methodology rests on seller-side diversion ratios derived from unreliable survey data and untested assumptions. The survey upon which he relies was designed after this litigation commenced, was framed in a manner that artificially constrained seller choices to marketplace platforms, and has not been validated for representativeness across the approximately 227,000 combined active sellers on the CascadeMarket and Pinnacle platforms. Moreover, Dr. Keenan's upward pricing pressure ("UPP") model fails entirely to account for the pervasive multi-homing behavior of online sellers and ignores $340 million in documented, merger-specific efficiencies.
These are not peripheral quibbles that can be addressed on cross-examination. Dr. Keenan's flawed market definition is the foundation upon which his concentration and competitive effects opinions are constructed, and the unreliability of that foundation infects every conclusion in his expert report. Permitting the FTC to present Dr. Keenan's opinions to this Court at the preliminary injunction hearing — opinions that fail to meet the threshold requirements for admissibility under Rule 702 — would undermine the gatekeeping function that Daubert and its progeny impose.
The evidentiary hearing on the FTC's preliminary injunction motion is set for September 8, 2025. Timely resolution of this motion is necessary to ensure that the hearing proceeds on the basis of reliable evidence and that the parties can prepare accordingly.
I. BACKGROUND
A. The Transaction
On March 14, 2025, Cascade entered into an Agreement and Plan of Merger (the "Merger Agreement") to acquire 100% of the outstanding common stock of Pinnacle for $26.85 per share in cash, representing aggregate consideration of approximately $4.7 billion. The transaction is structured as a reverse triangular merger under Delaware General Corporation Law, pursuant to which a wholly owned acquisition subsidiary of Cascade will merge with and into Pinnacle, with Pinnacle surviving the merger as a wholly owned subsidiary of Cascade.
Cascade is a Delaware corporation headquartered in Seattle, Washington. It operates a diversified omnichannel retail business that includes approximately 340 brick-and-mortar stores located across 38 states, a direct-to-consumer e-commerce platform operating under the brand name CascadeDirect.com, and a third-party seller marketplace known as CascadeMarket, which Cascade launched in June 2019 as part of a broader strategic initiative to expand its digital retail footprint. In fiscal year 2024, CascadeMarket generated approximately $3.1 billion in gross merchandise value ("GMV") from approximately 42,000 active third-party sellers, yielding $527 million in marketplace services revenue at an approximate take rate of 17%.
Pinnacle is a Delaware corporation headquartered in Austin, Texas. Unlike Cascade, Pinnacle operates exclusively as an online marketplace platform, hosting approximately 185,000 active sellers who processed approximately $11.4 billion in GMV in fiscal year 2024. Pinnacle generated approximately $1.82 billion in net revenue during that period, reflecting an approximate take rate of 16%. Pinnacle does not operate any brick-and-mortar retail locations and does not maintain a material direct-to-consumer e-commerce channel separate from its marketplace.
The financing for the proposed acquisition consists of approximately $1.9 billion in available cash on hand from Cascade's balance sheet and a $2.8 billion senior secured term loan B facility from Whitmore National Bank, N.A., as administrative agent and lead arranger. The commitment letter for the term loan facility was executed on March 12, 2025, two days before the execution of the Merger Agreement. The parties filed their premerger notification under the Hart-Scott-Rodino Antitrust Improvements Act of 1976, 15 U.S.C. § 18a, on April 7, 2025, and the Commission issued a Second Request for additional information and documentary material on May 9, 2025.
B. The FTC's Challenge
On June 20, 2025, the Commission — by a vote of 3-2 — authorized the filing of an administrative complaint (FTC Docket No. 9438) challenging the proposed transaction under Section 7 of the Clayton Act, 15 U.S.C. § 18. Three days later, on June 23, 2025, the FTC filed its complaint in this Court seeking a preliminary injunction under Section 13(b) of the FTC Act, 15 U.S.C. § 53(b), to enjoin the consummation of the merger pending the completion of administrative proceedings before the Commission.
The FTC alleges that the relevant product market in which to assess the competitive effects of the merger is "online marketplace platform services for third-party sellers in the United States." Within that proposed market, the FTC alleges the following shares, measured by GMV flowing through each platform's third-party marketplace:
Based on these market shares, the FTC alleges that the proposed acquisition would increase market concentration from a pre-merger HHI of 2,361 to a post-merger HHI of 2,856, representing an increase (delta) of 495. The Commission contends that these concentration levels, coupled with Dr. Keenan's competitive effects analysis, establish a presumption that the merger is likely to substantially lessen competition in the alleged relevant market.
C. Dr. Keenan's Engagement and Opinions
The FTC retained Dr. Wallace Keenan of Thornfield Research Group, located at 1515 Wilson Boulevard, Suite 300, Arlington, Virginia 22209, as its testifying economic expert in this matter. Dr. Keenan submitted his expert report on August 18, 2025, in accordance with the scheduling order entered by this Court. In his report, Dr. Keenan offers opinions on three principal topics: (1) the definition of the relevant product market; (2) market concentration as measured by HHI within the proposed market; and (3) predicted competitive effects of the merger, including an analysis of upward pricing pressure on seller take rates based on seller-side diversion ratios.
Dr. Keenan's opinions rest heavily on a hypothetical monopolist test — the SSNIP test — applied exclusively to third-party marketplace platform services. His competitive effects analysis is derived from seller-side diversion ratios that purport to measure the competitive closeness of CascadeMarket and Pinnacle as substitutes for third-party sellers. As detailed below, each of these analytical pillars suffers from fundamental methodological deficiencies that render Dr. Keenan's opinions unreliable and inadmissible under Rule 702 and Daubert.
II. LEGAL STANDARD
A. Federal Rule of Evidence 702
Federal Rule of Evidence 702, as amended effective December 1, 2023, provides:
A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that:
(a) 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;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product of reliable principles and methods; and
(d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.
Fed. R. Evid. 702. The 2023 amendment clarified the preponderance-of-the-evidence standard that governs the admissibility determination, emphasizing that "the proponent" — here, the FTC — "bears the burden of establishing" each element of admissibility. See Fed. R. Evid. 702 advisory committee's note to 2023 amendment ("The amendment clarifies that the preponderance standard applies to the Rule 702(d) determination as well, and that an opinion is not reliable simply because the expert says it is."). This amendment was specifically intended to address the concern that some courts were too readily admitting expert testimony by conflating the admissibility inquiry with the assessment of the weight or credibility of the testimony.
B. Daubert and Its Progeny
In Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589 (1993), the Supreme Court held that Federal Rule of Evidence 702 imposes upon the trial court a "gatekeeping" obligation to ensure that expert testimony is both relevant and reliable before it may be admitted. The Court identified several non-exclusive factors that may bear on the reliability of expert testimony, including: (1) whether the theory or technique can be and has been tested; (2) whether it has been subjected to peer review and publication; (3) the known or potential rate of error; and (4) general acceptance within the relevant scientific or professional community. Id. at 593-94. These factors are "meant to be helpful, not definitive," and courts are afforded "considerable leeway in deciding in a particular case how to go about determining whether particular expert testimony is reliable." Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999).
In General Electric Co. v. Joiner, 522 U.S. 136, 146 (1997), the Court emphasized that the gatekeeping function requires courts to examine whether there is "too great an analytical gap between the data and the opinion proffered." An expert opinion that is connected to existing data or methodology "only by the ipse dixit of the expert" is insufficient to satisfy the reliability standard. Id.
In Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), the Court extended the Daubert gatekeeping obligation to all forms of expert testimony, including non-scientific testimony based on "technical" or "other specialized" knowledge. Id. at 141. This holding is of particular relevance here, as Dr. Keenan's testimony rests on economic methodology rather than traditional scientific experimentation, but is no less subject to the Court's gatekeeping scrutiny.
C. Application to Economic Expert Testimony in Antitrust Cases
Courts presiding over merger challenges routinely exercise their gatekeeping function under Rule 702 and Daubert to exclude or limit the testimony of economic experts whose methodologies are unreliable or whose opinions are not grounded in a sound application of economic principles to the facts of the case. See, e.g., FTC v. Sysco Corp., 113 F. Supp. 3d 1, 25-26 (D.D.C. 2015) (evaluating reliability of economic expert's market definition analysis); United States v. H&R Pinnacle Holdings, 833 F. Supp. 2d 36, 72-73 (D.D.C. 2011) (same); In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 323 (3d Cir. 2008) (holding that a trial court must "resolve all Daubert issues before certifying a class," including the reliability of economic expert opinions on market definition).
Particular scrutiny is warranted where, as here, the expert's market definition is the keystone of the entire competitive effects analysis. Errors or unreliable assumptions embedded in the market definition propagate through the concentration analysis and into the competitive effects conclusions, amplifying the impact of any initial methodological flaw. See ABA Section of Antitrust Law, Mergers and Acquisitions 167 (4th ed. 2016) ("Market definition is often the central battleground in merger litigation because it determines market shares, concentration levels, and the identity of market participants."). As the Ninth Circuit has instructed, expert testimony that "offers no reliable basis on which to analyze" the economic question at issue is properly excluded. Daubert v. Merrell Dow Pharm., Inc., 43 F.3d 1311, 1319 (9th Cir. 1995) (on remand); see also Mukhtar v. Cal. State Univ., Hayward, 299 F.3d 1053, 1066 (9th Cir. 2002) ("[A]n expert who supplies nothing but a bottom line supplies nothing of value to the judicial process.").
III. ARGUMENT
A. Dr. Keenan's Market Definition Analysis Is Unreliable Because It Employs a Flawed SSNIP Test
Dr. Keenan defines the relevant product market as "online marketplace platform services for third-party sellers in the United States." This narrow market definition is the product of a hypothetical monopolist test that is fundamentally flawed in its design and execution. Specifically, Dr. Keenan's SSNIP analysis systematically ignores supply-side substitution and cross-platform competition, artificially constraining the relevant market and ensuring a result that supports the FTC's enforcement theory. These are not minor methodological disagreements; they are foundational errors that render Dr. Keenan's market definition unreliable and inadmissible.
1. The Hypothetical Monopolist / SSNIP Framework Requires Consideration of All Competitive Alternatives. The hypothetical monopolist test, also known as the SSNIP (Small but Significant and Non-transitory Increase in Price) test, is the standard analytical framework for defining relevant antitrust markets. Under this test, the analyst begins with a candidate market consisting of the products (or services) of the merging firms and asks whether a hypothetical monopolist of all products in the candidate market could profitably impose a small but significant and non-transitory increase in price — typically 5% to 10% — over prevailing levels. If a sufficient number of customers (or, on the sell side, suppliers) would switch to alternative products outside the candidate market in response to such a price increase, making it unprofitable, the candidate market must be broadened to include the next-best substitute. The process is repeated iteratively until the hypothetical monopolist could profitably sustain the price increase. See U.S. Dep't of Justice & FTC, Horizontal Merger Guidelines § 4.1.1 (2010); FTC & U.S. Dep't of Justice, Merger Guidelines § 4.3 (2023).
Critically, a properly conducted SSNIP test must account for both demand-side and supply-side substitution. Demand-side substitution refers to the willingness of customers (here, sellers) to switch to competing products or services in response to a price increase. Supply-side substitution refers to the ability of firms not currently participating in the candidate market to rapidly reposition their offerings to enter the market in response to a price increase, thereby constraining the hypothetical monopolist's ability to raise prices profitably. Both the 2010 and 2023 Horizontal Merger Guidelines expressly recognize that supply-side substitution can discipline prices just as effectively as demand-side switching and should be incorporated into the market definition analysis. See 2010 Guidelines § 5.1 ("The Agencies consider the competitive significance of possible supply responses in defining markets and measuring market shares."); 2023 Guidelines § 4.3.C (recognizing that "rapid supply-side entry" may constrain market power).
The failure to account for either dimension of substitution in a SSNIP test produces a candidate market that is artificially narrow and therefore unreliable. An artificially narrow market, by definition, overstates the market power of the merging firms and exaggerates the competitive significance of the proposed transaction.
2. Dr. Keenan's SSNIP Test Ignores Supply-Side Substitution. Dr. Keenan defined the candidate market as "online marketplace platform services for third-party sellers" and conducted his hypothetical monopolist test by examining only whether third-party sellers would switch to other dedicated marketplace platforms in response to a take-rate increase. Keenan Report ¶¶ 42-58. His analysis entirely omitted the supply side — that is, the demonstrated ability and economic incentive of firms operating outside the candidate market to rapidly offer marketplace-type services to third-party sellers in response to a supra-competitive increase in take rates.
The evidence of supply-side substitution is substantial and readily available. At least three categories of firms are positioned to offer, and in several cases have already begun offering, marketplace-type services to third-party sellers:
Direct-to-consumer e-commerce platforms. Cascade itself operates CascadeDirect.com, a direct-to-consumer e-commerce platform that generated approximately $40.6 billion in sales in FY2024. The very existence of CascadeMarket — launched in 2019 by repurposing Cascade's existing e-commerce infrastructure — demonstrates that a direct-to-consumer platform can be repositioned to serve third-party sellers at relatively low marginal cost and within a short time frame. Other major e-commerce retailers, including at least three firms with revenues exceeding $5 billion, operate or are developing similar third-party seller capabilities. Dr. Keenan's SSNIP test entirely ignores these competitive alternatives.
Social commerce platforms. The emergence of social commerce represents a rapidly growing competitive alternative to dedicated marketplace platforms. Brevity Social Commerce, launched in 2023, enables third-party sellers to list and sell products directly through integrated social media storefronts. Social commerce platforms generated an estimated $67 billion in U.S. GMV in 2024, a figure projected to grow at a compound annual rate of approximately 25% through 2028. These platforms compete directly with marketplace platforms for third-party seller volume, yet Dr. Keenan excluded them from his analysis without providing any economic justification.
Hybrid retail models. Trove Marketplace, launched in 2024, represents a new category of hybrid retail platform that combines elements of traditional e-commerce, social commerce, and marketplace functionality. Trove and similar platforms allow third-party sellers to access consumers through multiple channels within a single integrated environment. The entry of Trove and comparable platforms demonstrates that the barriers to supply-side entry into marketplace services for third-party sellers are low and that repositioning by adjacent competitors is both feasible and occurring in the marketplace today.
By ignoring supply-side substitution from these and other competitive alternatives, Dr. Keenan artificially constrained the set of products and services included in his candidate market and ensured that his SSNIP test would yield a narrow market limited to dedicated marketplace platforms. This is a results-driven methodological choice, not an analytically sound application of the hypothetical monopolist framework. It renders Dr. Keenan's market definition unreliable under Rule 702 and Daubert.
3. Dr. Keenan's SSNIP Test Ignores Cross-Platform Competition. Dr. Keenan's SSNIP test is further undermined by his failure to account for the pervasive cross-platform competition that characterizes the e-commerce sector. Online sellers routinely engage in "multi-homing" — that is, they sell their products across multiple platforms simultaneously, distributing their sales volume among marketplace platforms, direct-to-consumer storefronts, social media commerce channels, and hybrid platforms.
The evidence of multi-homing is extensive and directly relevant to the SSNIP analysis. Internal Cascade documents produced in discovery, as well as internal Pinnacle documents, demonstrate that a substantial majority of active sellers on both CascadeMarket and Pinnacle simultaneously maintain direct-to-consumer e-commerce storefronts, sell through one or more social media commerce channels, or utilize hybrid platforms. Publicly available industry surveys corroborate this finding. For example, the 2024 E-Commerce Seller Survey published by the National Retail Federation found that 73% of sellers active on at least one online marketplace also sell through at least one non-marketplace channel. Similarly, Pinnacle's own internal data, as reflected in its FY2024 10-K filing, notes that "a significant portion of our active sellers maintain selling presences on multiple platforms, including direct-to-consumer e-commerce sites and social commerce channels."
A proper SSNIP analysis must account for this cross-platform competition because multi-homing means that a take-rate increase on one platform causes marginal sellers to shift volume — not necessarily their entire business — to competing channels. The relevant constraint on a hypothetical monopolist's ability to raise take rates profitably is not limited to dedicated marketplace platforms; it includes every alternative channel through which a seller can reach consumers and generate sales. A seller who maintains a storefront on Pinnacle, a direct-to-consumer website, and a Brevity Social Commerce shop does not need to abandon Pinnacle entirely in response to a take-rate increase. Instead, the seller can shift marginal advertising spend, inventory allocation, and promotional activity toward competing channels — a form of competitive discipline that is invisible to Dr. Keenan's marketplace-only SSNIP test.
Dr. Keenan treated marketplace platforms as though they exist in a competitive vacuum, isolated from the broader $757 billion U.S. e-commerce ecosystem (2024 estimate, U.S. Census Bureau). This assumption is contradicted by overwhelming evidence of cross-platform competition and multi-homing. Cascade's expert, Dr. Priya Ramanathan of Brindoval Analytics LLC, demonstrates in her analysis that when cross-platform substitution is properly accounted for, the relevant market encompasses all e-commerce retail platforms competing for seller volume. In that properly defined market, the combined entity's share is approximately 7.3% of total U.S. e-commerce GMV — a share level at which no plausible theory of competitive harm can be sustained.
4. The Analytical Gap Between Dr. Keenan's Data and His Market Definition Conclusion. Even accepting Dr. Keenan's own analytical framework arguendo, the data he assembled do not support the narrow market definition he reached. This analytical gap is precisely the kind of disconnect between data and opinion that the Supreme Court identified in Joiner as grounds for exclusion.
Dr. Keenan's own survey data reveal that 34% of sellers surveyed indicated that they would consider shifting sales volume to non-marketplace channels — including direct-to-consumer websites and social commerce platforms — in response to a take-rate increase of the magnitude specified in his hypothetical monopolist test. Keenan Report, Appendix C, Table 5. Under standard SSNIP methodology, this level of switching would ordinarily compel the analyst to expand the candidate market to include those non-marketplace channels, because the diversion of even a fraction of this volume could render the hypothetical monopolist's price increase unprofitable.
Rather than following the logic of his own data, Dr. Keenan disregarded this evidence by applying an ad hoc "critical loss" threshold that is inconsistent with his own stated methodology. Specifically, Dr. Keenan asserted in his report that sellers who indicated a willingness to shift volume to non-marketplace channels should be disregarded because, in his judgment, those sellers were "unlikely to complete a full transition" away from marketplace platforms. Keenan Report ¶ 54. But the SSNIP test does not require a "full transition"; it requires only that sufficient volume shift to make the price increase unprofitable. By imposing a full-transition requirement that has no basis in accepted economic methodology, Dr. Keenan effectively moved the goalposts to ensure that his SSNIP test produced the narrow market definition the FTC's theory demands.
This is the paradigmatic "analytical gap" between data and conclusion that the Joiner Court warned against. Dr. Keenan's data point in one direction — toward a broader market — but his conclusion points in another. The gap is bridged not by rigorous economic analysis but by the ipse dixit of the expert. Under Daubert and Rule 702, this gap warrants exclusion.
B. Dr. Keenan's HHI Concentration Analysis Is Unreliable Because It Is Derived From the Flawed Market Definition
1. The HHI Calculations Are Only as Reliable as the Underlying Market. Dr. Keenan calculates a pre-merger HHI of 2,361, a post-merger HHI of 2,856, and an HHI increase (delta) of 495. The HHI delta is derived mechanically from the market shares of CascadeMarket and Pinnacle within the proposed market: 2 × 8.2 × 30.1 = 493.64, which the FTC rounds to 495 in its Complaint at ¶ 47.
These figures are derived exclusively from market shares measured within the FTC's proposed narrow market of "online marketplace platform services for third-party sellers in the United States." The share figures — Cascade 8.2%, Pinnacle 30.1%, Vertex Commerce 35.7%, Nomad Shops 14.6%, and all others 11.4% — are meaningful only within the confines of that particular market definition. If the market definition is unreliable, as Cascade has demonstrated above, then the HHI calculations constructed upon it are necessarily unreliable as well.
The principle is straightforward: "garbage in, garbage out." A mathematically precise HHI computation does not cure a fundamentally flawed market definition input. The HHI is merely a summary statistic that describes the distribution of market shares within a defined market. If the market is defined too narrowly — as it is here — then the shares allocated to each firm will be inflated relative to shares that would prevail in a properly defined market, and the resulting HHI will overstate the degree of concentration. The precision of the arithmetic is irrelevant if the inputs to the arithmetic are unreliable. See FTC v. Sanford Health, 926 F.3d 959, 965 (8th Cir. 2019) (recognizing that market definition determines the universe of competitors whose shares are measured).
2. A Properly Defined Market Yields Dramatically Different Concentration Results. The magnitude of the distortion introduced by Dr. Keenan's narrow market definition is readily apparent when concentration is measured in a properly defined market. In a market defined as all U.S. e-commerce retail — which, as set forth above, encompasses the full range of competitive alternatives available to third-party sellers — total GMV in 2024 was approximately $757 billion (U.S. Census Bureau estimate).
Within that broader market, the combined entity's marketplace GMV of approximately $14.5 billion ($3.1 billion from CascadeMarket plus $11.4 billion from Pinnacle) represents approximately 1.9% of the market. Even including Cascade's total e-commerce GMV — approximately $43.7 billion, including both direct-to-consumer sales through CascadeDirect.com and third-party marketplace sales through CascadeMarket — the combined entity's share is approximately 7.3% ($55.1 billion divided by $757 billion).
At these share levels, the post-merger HHI would fall far below the thresholds at which any version of the Horizontal Merger Guidelines identifies competitive concern. Under both the 2010 and 2023 Guidelines, mergers that result in post-merger HHI levels below 1,500 are generally considered unlikely to raise competitive concerns, and mergers resulting in HHI levels between 1,500 and 2,500 are considered potentially concerning only if the delta exceeds certain thresholds. A combined share of 7.3% in a market with numerous large competitors would not approach these thresholds.
Dr. Keenan never analyzed concentration in the broader market. He did not compute HHI levels for any market definition other than the narrow one the FTC's theory requires. This omission further demonstrates the outcome-determinative nature of his narrow market definition: Dr. Keenan's entire analytical framework is designed to produce concentration figures that support a presumption of illegality, rather than to conduct an objective assessment of the competitive significance of the proposed transaction.
3. Dr. Keenan's Market Share Measurement Methodology Contains Additional Flaws. Even within the narrow market Dr. Keenan has defined, his methodology for measuring market shares contains additional errors that compound the market definition problem and render his HHI calculations even less reliable.
First, Dr. Keenan measured shares by GMV flowing through the seller side of each marketplace platform, but failed to account for the two-sided nature of the platform. Online marketplace platforms are classic two-sided transaction platforms: they serve both sellers (who list products and pay take rates) and consumers (who purchase products). The Supreme Court's analysis in Ohio v. Halcyon Corporation, 138 S. Ct. 2274 (2018), instructs that both sides of a two-sided transaction platform must be considered in assessing competitive effects. While Amex addressed the substantive antitrust analysis rather than Rule 702 admissibility, its recognition that both sides of a platform matter is directly relevant to the reliability of Dr. Keenan's one-sided market share methodology. A market share analysis that accounts for seller-side GMV but ignores consumer-side transaction volume and competitive dynamics produces a distorted picture of competitive significance.
Second, Dr. Keenan excluded from his share calculations all transaction volume that flows through non-marketplace e-commerce channels operated by the same firms whose marketplace shares he measured. Most notably, he measured Cascade's share using only the $3.1 billion in GMV from CascadeMarket, while excluding the approximately $40.6 billion in direct-to-consumer sales through CascadeDirect.com. This exclusion is analytically problematic because CascadeDirect.com competes directly with marketplace platforms for seller volume (through vendor relationships and wholesale purchasing) and for consumer purchasing. By measuring Cascade's share as only 8.2% rather than a figure reflecting its total e-commerce volume, Dr. Keenan's methodology paradoxically understates Cascade's competitive significance in some respects while simultaneously overstating the merger's impact on concentration by confining the analysis to a narrow market in which the combination appears more consequential.
These measurement errors compound the market definition error described in Section III.A and render the HHI figures presented in Dr. Keenan's report unreliable as a foundation for expert testimony.
C. Dr. Keenan's Competitive Effects Analysis Relies on Unreliable Seller-Side Diversion Ratios
1. Description of Dr. Keenan's Diversion Ratio Methodology. Beyond his market definition and concentration opinions, Dr. Keenan offers opinions regarding the predicted competitive effects of the merger. The centerpiece of his competitive effects analysis is a set of seller-side diversion ratios purporting to demonstrate that CascadeMarket and Pinnacle are "close competitors" — that is, that a significant share of the sellers who would leave one platform in response to a take-rate increase would switch to the other platform. Keenan Report ¶¶ 72-91.
Dr. Keenan used these diversion ratios as inputs to an upward pricing pressure ("UPP") model. The UPP framework estimates the incentive of the merged firm to raise price (here, take rates) by comparing the value of sales recaptured through the merger partner (measured by the diversion ratio multiplied by the margin earned on diverted sales) against any countervailing downward pressure from merger-specific efficiencies. Dr. Keenan's UPP analysis purports to show that the merged entity would have an incentive to raise take rates for third-party sellers on both platforms, resulting in competitive harm. Keenan Report ¶¶ 92-108.
The reliability of Dr. Keenan's competitive effects analysis depends entirely on the reliability of his diversion ratio estimates, which in turn depend on the reliability of the survey data from which they are derived. As set forth below, both the survey data and the diversion ratio methodology are fatally flawed.
2. The Underlying Survey Data Are Unreliable. Dr. Keenan's seller-side diversion ratios are based on a survey of third-party sellers conducted by Thornfield Research Group. This survey suffers from fundamental design flaws that render it unreliable as a foundation for expert testimony under Rule 702 and Daubert.
Litigation-driven framing. The survey was designed and administered after the FTC filed its administrative complaint on June 20, 2025. Keenan Report, Appendix D (survey methodology) (survey fielded July 7-21, 2025). While the timing of a survey does not automatically disqualify it, the post-litigation design of the survey raises serious concerns that it was framed to generate results supportive of the FTC's enforcement theory rather than to objectively measure seller switching behavior. Courts have recognized that surveys designed in the context of pending litigation warrant heightened scrutiny. See Shari Scharlin, Survey Evidence in Antitrust Proceedings, 23 Antitrust L.J. 355 (2019).
Artificially constrained choice set. The survey presented sellers with a hypothetical question about their switching behavior in response to a take-rate increase, but the hypothetical was framed in a manner that limited seller response options to other marketplace platforms. Specifically, the question asked: "If [platform name] increased its take rate by [X]%, to which online marketplace platform would you most likely shift your sales volume?" Keenan Report, Appendix D, Survey Instrument, Q12. By embedding the phrase "online marketplace platform" in the question itself, the survey effectively excluded direct-to-consumer and social commerce alternatives from the choice set. Sellers who would have indicated a preference for shifting volume to their own website, to a social commerce channel, or to a hybrid platform were not given the opportunity to express that preference. The resulting diversion ratios are therefore biased upward toward marketplace-to-marketplace diversion, precisely the outcome the FTC's theory requires.
Inadequate validation for representativeness. The combined active seller base on CascadeMarket and Pinnacle totals approximately 227,000 sellers (42,000 on CascadeMarket and 185,000 on Pinnacle). Dr. Keenan does not disclose the total number of survey responses or the response rate in the body of his report, noting only in Appendix D that the survey achieved "a sufficient number of responses for statistical analysis." Keenan Report, Appendix D, at 3. Nor does Dr. Keenan demonstrate that the survey sample is representative of the seller population along critical dimensions, including seller size (by GMV), product category, tenure on the platform, and degree of multi-homing. Without adequate validation for representativeness, the survey cannot be relied upon to generate diversion ratio estimates that are generalizable to the seller population as a whole.
An expert opinion constructed on the foundation of unreliable survey data is itself unreliable under Rule 702 and Daubert. Where, as here, the survey instrument contains design flaws that bias the results in a predictable direction, the survey cannot serve as a reliable basis for the expert's conclusions.
3. The Diversion Ratio Model Ignores Multi-Homing. Even if Dr. Keenan's survey data were reliable — and they are not — his diversion ratio framework would still be analytically unsound because it fails to account for the multi-homing behavior of online sellers.
As discussed in Section III.A.3 above, a substantial majority of sellers on both CascadeMarket and Pinnacle sell simultaneously on multiple platforms, including marketplace platforms, direct-to-consumer websites, and social commerce channels. Multi-homing is not an incidental feature of the e-commerce landscape; it is a defining characteristic that fundamentally shapes seller behavior and the competitive dynamics of the industry.
Multi-homing undermines the analytical coherence of the diversion ratio framework in several ways. Most fundamentally, if a seller already sells on both CascadeMarket and Pinnacle — as many sellers do — the concept of "diversion" from one platform to the other is analytically incoherent. A seller who is present on both platforms cannot be "diverted" to a platform where it already operates. What the seller can do, in response to a take-rate increase on one platform, is shift marginal volume toward the other platform (or toward other channels), but this marginal volume shift is conceptually and quantitatively distinct from the wholesale diversion of a seller's business that Dr. Keenan's model assumes.
Dr. Keenan's diversion ratio model fails to distinguish between diversion of marginal volume and diversion of an entire seller relationship. By treating every seller as though it either sells on CascadeMarket or Pinnacle but not both, the model overstates the degree to which a take-rate increase on one platform would benefit the other, leading to inflated diversion ratio estimates that overstate the merged firm's incentive to raise take rates.
Cascade's rebuttal expert, Dr. Priya Ramanathan of Brindoval Analytics LLC, presents an alternative econometric model in her expert report that explicitly accounts for seller multi-homing. Dr. Ramanathan's analysis demonstrates that when multi-homing rates are properly incorporated into the competitive effects framework, the predicted diversion from CascadeMarket to Pinnacle (and vice versa) drops substantially, rendering the FTC's unilateral effects theory implausible. Dr. Ramanathan's rebuttal report, due August 29, 2025 under the Court's scheduling order, will be submitted in advance of the evidentiary hearing and will provide the Court with a methodologically sound alternative to Dr. Keenan's flawed analysis.
4. The UPP Analysis Fails to Account for Efficiencies. Even setting aside the deficiencies in Dr. Keenan's diversion ratios, his upward pricing pressure analysis is independently unreliable because it fails to account for cognizable, merger-specific efficiencies.
A properly specified UPP model incorporates merger-specific efficiencies as a downward offset against the upward pricing pressure generated by the internalization of competitive externalities between the merging firms. See Joseph Farrell & Carl Shapiro, Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition, 10 B.E. J. Theoretical Econ. 1, 9-10 (2010) (describing the role of efficiencies as a "default efficiency" offset in the UPP framework). The Horizontal Merger Guidelines similarly recognize that "cognizable efficiencies" — those that are merger-specific, verifiable, and not derived from anticompetitive reductions in output or service — should be credited in the competitive effects analysis. 2010 Guidelines § 10; 2023 Guidelines § 3.4.
Cascade has documented $340 million in annual merger-specific efficiencies projected to be achieved by Year 3 following consummation of the transaction. These efficiencies include:
• Logistics and fulfillment cost savings: $195 million. Integration of Pinnacle's third-party sellers into Cascade's existing 22-warehouse fulfillment network, which currently serves CascadeMarket sellers as well as Cascade's direct-to-consumer business. Cascade's fulfillment infrastructure has significant excess capacity that can absorb Pinnacle's seller volume at marginal cost. The resulting efficiencies include reduced per-unit shipping costs, consolidation of inbound logistics, and elimination of redundant fulfillment center leases currently maintained by Pinnacle's third-party logistics partners.
• Technology and platform synergies: $145 million. Consolidation of duplicative technology platforms, including seller management tools, search and discovery algorithms, payment processing systems, and customer service infrastructure. Cascade's investment of over $800 million in its e-commerce technology stack since 2019 provides a scalable foundation that can be extended to Pinnacle's seller base without proportional incremental cost.
These efficiencies are merger-specific (they cannot be achieved absent the transaction), verifiable (they are supported by detailed integration planning documents and third-party validation by Cascade's financial advisor), and would be passed through to sellers in the form of lower costs and improved services.
Dr. Keenan's UPP analysis either ignores these efficiencies entirely or dismisses them as "non-cognizable" without conducting any independent verification analysis. Keenan Report ¶¶ 105-108. Dr. Keenan does not examine Cascade's integration planning documents, does not interview Cascade's operations team, and does not apply any recognized methodology for evaluating whether claimed efficiencies are merger-specific and verifiable. He simply asserts, in a conclusory fashion, that "the claimed efficiencies are speculative and unverified." Keenan Report ¶ 107. This is not economic analysis; it is advocacy. A UPP model that ignores one of its essential inputs — the efficiency offset — is incomplete and unreliable, and the resulting competitive effects opinions should be excluded.
D. The Flaws in Dr. Keenan's Analysis Are Not Matters of Weight But of Admissibility
Cascade anticipates that the FTC will respond to this motion by arguing that the deficiencies identified above go merely to the "weight, not the admissibility" of Dr. Keenan's testimony, and that they can be adequately addressed through vigorous cross-examination at the evidentiary hearing. This argument should be rejected.
The distinction between weight and admissibility is a meaningful one, but it does not operate to excuse fundamental methodological errors of the kind present in Dr. Keenan's analysis. A disagreement between competing experts about the correct value of an input parameter, or about the relative probative value of different data sources, is typically a matter of weight. But the wholesale omission of an analytically essential category of competitive alternatives (supply-side substitution and cross-platform competition) from a SSNIP test, the use of survey data with a biased choice set, and the failure to account for a defining feature of the competitive landscape (multi-homing) are not differences of degree — they are differences of kind. They reflect not a reasonable exercise of expert judgment but a systematic departure from accepted economic methodology.
Moreover, the errors in Dr. Keenan's analysis are not isolated. They are interconnected and mutually reinforcing. The narrow market definition produces inflated market shares, which produce inflated HHI levels, which create an appearance of high concentration that is then compounded by the inflated diversion ratios derived from the flawed survey. The entire chain of reasoning is built on an unreliable foundation, and removing any one link does not cure the deficiency because the remaining links are themselves infected by the same foundational errors.
The 2023 amendment to Rule 702 was specifically intended to correct the tendency of some courts to "admit expert testimony too readily" by conflating the admissibility determination with the assessment of weight. Fed. R. Evid. 702 advisory committee's note to 2023 amendment. The amendment reinforces that "the proponent must demonstrate to the court that it is more likely than not that the proffered testimony meets the admissibility requirements set out in the rule." Id. Here, the FTC cannot carry that burden, because the methodological flaws in Dr. Keenan's analysis are sufficiently fundamental to render his opinions unreliable under any reasonable application of Rule 702 and Daubert.
Under Daubert and its progeny, it is the Court's responsibility as gatekeeper to exclude expert testimony that fails to meet reliability thresholds. The gatekeeping function exists precisely to prevent unreliable evidence from being presented to the finder of fact, because once such evidence has been received, the damage cannot always be undone through cross-examination. As the Supreme Court recognized, "vigorous cross-examination" and "presentation of contrary evidence" are not adequate substitutes for the Court's independent assessment of reliability. Daubert, 509 U.S. at 596. Rather than leaving it to Cascade to "unring the bell" through cross-examination, the Court should exercise its gatekeeping authority to exclude Dr. Keenan's unreliable testimony.
E. In the Alternative, the Court Should Limit Dr. Keenan's Testimony
Should the Court decline to exclude Dr. Keenan's testimony in its entirety, Cascade respectfully requests, in the alternative, that the Court enter an order limiting the scope of Dr. Keenan's testimony at the September 8, 2025 evidentiary hearing as follows:
First, the Court should preclude Dr. Keenan from testifying to the conclusion that the relevant product market is limited to "online marketplace platform services for third-party sellers in the United States." This conclusion is the product of a SSNIP test that systematically excludes supply-side substitution and cross-platform competition, and it does not meet the reliability threshold required by Rule 702. Dr. Keenan should be permitted to testify regarding economic principles generally, but not to offer the specific market definition opinion that is the product of his flawed methodology.
Second, the Court should preclude Dr. Keenan from offering HHI calculations — specifically, his figures of 2,361 (pre-merger), 2,856 (post-merger), and 495 (delta) — based solely on market shares within the narrow proposed market. In the alternative, the Court should require Dr. Keenan to present alternative HHI calculations based on a properly defined broader market that accounts for all relevant competitive alternatives, so that the Court may assess concentration in its proper competitive context.
Third, the Court should preclude Dr. Keenan from offering competitive effects opinions derived from seller-side diversion ratios generated by the flawed survey methodology described above. Because the diversion ratios are the product of a survey with an artificially constrained choice set and inadequate validation for representativeness, and because the diversion ratio model fails to account for multi-homing, the competitive effects opinions built upon them do not satisfy the reliability requirements of Rule 702.
These limitations would permit the FTC to present whatever reliable economic testimony Dr. Keenan can offer, while preventing the Court from being presented with unreliable opinions that do not satisfy the threshold requirements for admissibility.
IV. CONCLUSION
For the foregoing reasons, Defendant Cascade Retail Holdings Inc. respectfully requests that this Court grant its motion and enter an order excluding the expert testimony of Dr. Wallace Keenan in its entirety or, in the alternative, limiting his testimony as specified in Section III.E above.
This motion is timely filed in advance of the evidentiary hearing on the FTC's motion for a preliminary injunction, which is set for September 8, 2025. Cascade respectfully submits that resolution of this motion prior to the hearing is necessary to ensure that the hearing proceeds on the basis of reliable evidence and that the parties and the Court can allocate hearing time accordingly.
Cascade respectfully requests oral argument on this motion if the Court deems it appropriate.
Respectfully submitted,
CALDERWOOD & HARKNESS LLP
By: ________
Howard Bellamy D.C. Bar No. 487231 Calderwood & Harkness LLP 1401 K Street NW, Suite 1200 Washington, D.C. 20005 Telephone: (202) 555-4100 Facsimile: (202) 555-4101
Email: hbellamy@calderwoodharkness.com
Aaron Liu Cal. Bar No. 321574 Calderwood & Harkness LLP 555 California Street, Suite 2700 San Francisco, CA 94104 Telephone: (415) 555-6200 Facsimile: (415) 555-6201
Email: aliu@calderwoodharkness.com
Counsel for Defendant Cascade Retail Holdings Inc.
Matter No.: 1038-00001
Dated: August 25, 2025
CERTIFICATE OF SERVICE
I hereby certify that on August 25, 2025, I caused a true and correct copy of Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan, together with all supporting papers, to be served upon the following counsel of record via the Court's CM/ECF electronic filing system, which will send notification of such filing to all registered participants:
Counsel for Plaintiff Federal Trade Commission:
Claire Dunmore, Lead Trial Attorney Samuel Ortega, Senior Trial Counsel Rebecca Yun, Staff Attorney Federal Trade Commission Bureau of Competition, Mergers IV Division 600 Pennsylvania Avenue NW Washington, D.C. 20580
Counsel for Co-Defendant Pinnacle Market Group Inc.:
Diana Whitfield Kenneth Oshiro Whitfield, Greer & Tanaka LLP 100 Congress Avenue, Suite 2200 Austin, TX 78701
Counsel for Proposed Intervenor Independent Online Sellers Alliance (IOSA):
Martina Galloway Galloway & Peck LLP 60 East 42nd Street, Suite 2600 New York, NY 10165
________________________________________
Howard Bellamy Calderwood & Harkness LLP
UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE
[PROPOSED] ORDER GRANTING DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN
This matter comes before the Court on Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan, filed August 25, 2025. The Court has considered the motion and all supporting papers, the opposition filed by Plaintiff Federal Trade Commission, and the reply filed by Defendant Cascade Retail Holdings Inc. Having considered the submissions of the parties, the relevant legal authorities, and the full record in this matter, the Court finds as follows:
IT IS HEREBY ORDERED that Defendant Cascade Retail Holdings Inc.'s Motion to Exclude or Limit the Expert Testimony of Dr. Wallace Keenan is GRANTED [or GRANTED IN PART].
IT IS FURTHER ORDERED that Dr. Wallace Keenan is precluded from offering the following testimony at the September 8, 2025 preliminary injunction evidentiary hearing:
________________________________________
________________________________________
________________________________________
________________________________________
SO ORDERED.
Dated: ________
________________________________________
THE HONORABLE RICHARD A. JOHANNSEN United States District Judge
FEDERAL TRADE COMMISSION, | Case No. 2:25-cv-01087-RAJ
Plaintiff, | Before the Honorable Richard A. Johannsen
v. | DEFENDANT CASCADE RETAIL HOLDINGS INC.'S MOTION TO EXCLUDE OR LIMIT THE EXPERT TESTIMONY OF DR. WALLACE KEENAN
CASCADE RETAIL HOLDINGS INC. and PINNACLE MARKET GROUP INC.,
Defendants. | Evidentiary Hearing: September 8, 2025
Firm | GMV (FY2024) | Market Share
Vertex Commerce Inc. | $13.5 billion | 35.7%
Pinnacle Market Group Inc. | $11.4 billion | 30.1%
Nomad Shops LLC | $5.5 billion | 14.6%
Cascade Retail Holdings Inc. (CascadeMarket) | $3.1 billion | 8.2%
All Others | $4.3 billion | 11.4%
Total | $37.8 billion | 100.0%
FEDERAL TRADE COMMISSION, | Case No. 2:25-cv-01087-RAJ
Plaintiff,
v.
CASCADE RETAIL HOLDINGS INC. and PINNACLE MARKET GROUP INC.,
Defendants.

UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE

INTRODUCTION

I. BACKGROUND

II. LEGAL STANDARD

III. ARGUMENT

IV. CONCLUSION

CALDERWOOD & HARKNESS LLP

CERTIFICATE OF SERVICE

UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT SEATTLE

SO ORDERED.

Record

Parties

Dates

Amounts