memo-license-back-market-survey.docx

1025-00008 · Quorum Insurance Holdings · IP & Derived Data Analysis

memo-license-back-market-survey.docx DOCUMENT M&A / Corporate

37 / 76 ← → ✕

MEMORANDUM

TO: Wendy Schultz, Associate, intellectual Property, Calderwood & Harkness LLP Leonard Cross, Partner, intellectual Property, Calderwood & Harkness LLP

FROM: Omar Siddiqui, Associate, intellectual Property, Calderwood & Harkness LLP

DATE: April 2023

RE: Market Survey — License-Back Provisions in Enterprise Data-Licensing Transactions

MATTER: Quorum Insurance Holdings Corp. — Meridian Credit intelligence, Inc. Data License Negotiation; C&H Matter No. 1025-00008

CC: Patricia Heng, Senior Counsel, Technology Transactions, Calderwood & Harkness LLP

Right-click to update Table of Contents

I. Executive Summary

This memorandum presents the results of a market survey of license-back provisions in enterprise data-licensing transactions, prepared in connection with Calderwood & Harkness LLP's representation of Quorum Insurance Holdings Corp. ("Quorum") in its ongoing negotiation of the proposed Master Data License Agreement ("MDLA") with Meridian Credit intelligence, Inc. ("Meridian"). The survey focuses specifically on arrangements in which a licensee creates derived data — including blended risk scores, propensity models, enriched datasets, and other derivative analytics outputs — from licensed third-party data, and the data licensor claims rights to such derived outputs through license-back provisions, commercialization restrictions, or revenue-sharing mechanisms.

The impetus for this survey is Meridian's demand, communicated through its outside counsel at Kessler Pruitt & Ashe LLP (Graham Whitley) and its internal negotiating team led by Arthur Pennington (General Counsel) and Keith Marlow (SVP, Data Licensing & Partnerships), for a license-back and revenue-sharing package relating to Derived Data (as that term is defined in the current draft of the MDLA). Meridian's proposed terms consist of three components:

(a) Perpetual, Royalty-Free License-Back. Meridian would receive a perpetual, royalty-free, non-exclusive license-back on all Derived Data created by Quorum using Meridian's licensed data as an input. This license-back would permit Meridian to use all such Derived Data for its own internal and commercial purposes and would survive termination of the MDLA.

(b) Commercialization Gating with Sole Discretion. Quorum would be prohibited from selling or sublicensing Derived Data to third parties unless the Derived Data has been "sufficiently transformed" such that no Meridian raw data can be reverse-engineered from the output. Critically, Meridian would retain sole discretion to determine whether any particular output meets the "sufficient transformation" standard.

(c) 15% Revenue Share. If Quorum receives Meridian's approval to commercialize Derived Data to third parties, Meridian would receive 15% of net revenues attributable to such Derived Data.

Overall Conclusion. Based on the market survey detailed below, Meridian's proposed license-back package — combining a perpetual royalty-free license-back, sole-discretion transformation gatekeeping, and a 15% revenue share — is materially outside market norms for enterprise data-licensing transactions, both when each element is considered individually and, most strikingly, when the elements are considered in combination. No surveyed agreement combined all three of these mechanisms, and several individual components of Meridian's proposal (particularly the sole-discretion gatekeeping mechanism and the 15% revenue-share rate) have no precedent in the agreements we reviewed.

Specific Strategic Concern. The license-back proposal poses a direct threat to Quorum's strategic investment in QuorumScore™, a proprietary predictive underwriting model (USPTO trademark application pending, Serial No. 97/482,116) that blends Meridian credit data with Quorum's first-party claims, policy, and telematics data. QuorumScore™ is central to Quorum's planned B2B licensing revenue line, which is projected to generate $40–50 million in annual revenues by 2027. Under Meridian's proposed terms, Quorum's ability to commercialize QuorumScore™ outputs to third parties would be subject to Meridian's sole-discretion approval, and even if approved, would be burdened by a 15% revenue share that would transfer $6–7.5 million annually to Meridian — a material sum that would fundamentally alter the economics of the analytics strategy. Moreover, Meridian would receive a perpetual, royalty-free license to use QuorumScore™ outputs for its own commercial purposes, including in markets where Meridian competes directly with Quorum's planned downstream analytics offerings.

This memorandum sets forth the detailed findings of the market survey, provides a comparative analysis of Meridian's proposal against observed market practice, and offers recommended negotiating positions for Quorum.

II. Scope and Methodology

A. Scope of Survey

This survey examined approximately thirty-five (35) enterprise data-licensing agreements spanning the credit-data, financial-data, consumer-data, and analytics sectors. The agreements were sourced from a combination of publicly available materials and proprietary research, including:

• Agreements filed as exhibits to SEC registration statements, annual reports, and current reports (Forms S-1, 10-K, 10-Q, and 8-K), as well as agreements produced in connection with regulatory submissions to state insurance departments and federal banking regulators;

• Proprietary databases maintained by Calderwood & Harkness LLP's Technology Transactions and intellectual Property practice groups, including transaction databases reflecting the firm's representation of clients in comparable data-licensing transactions;

• Reported litigation involving disputes over derived-data ownership, license-back provisions, and revenue-sharing obligations in data-licensing arrangements, including court opinions and publicly filed pleadings that contained or described relevant contractual terms; and

• Industry commentary and guidance materials published by trade associations (including the Data & Marketing Association and the International Association of Privacy Professionals), law firm client memoranda, and academic literature addressing market practice in data licensing.

The survey focused specifically on agreements in which the licensee uses licensed data as an input to create derivative analytics products, models, scores, or enriched datasets — i.e., transactions in which the derived-data ownership question is commercially material. We excluded from the survey two categories of agreements that present different commercial dynamics and are not directly comparable to the Quorum-Meridian transaction:

• Pure reseller/redistribution licenses, in which the licensee does not create derived outputs but instead redistributes the licensor's data (or lightly formatted versions thereof) to end users. In such arrangements, the licensor typically retains ownership of all data and the licensee acts as a distribution channel, with economics structured as a resale margin rather than a license fee. These arrangements do not raise the derived-data ownership issues central to the Quorum-Meridian negotiation.

• Intra-group data-sharing arrangements, in which data flows between affiliated entities within a single corporate group. These arrangements are governed by different considerations (including tax, regulatory, and intercompany pricing requirements) and do not reflect arm's-length market terms.

The time period covered by the survey encompasses agreements executed or publicly disclosed between 2017 and 2023. This six-year window was selected to capture current market practice while providing sufficient historical depth to identify trends and variations.

B. Definitions Used

For purposes of this memorandum, the following terms are used consistently:

"License-Back" refers to a provision in a data-licensing agreement that grants the data licensor a license to use data, models, analytics, scores, or other outputs created by the licensee using the licensor's data as an input. License-back provisions vary widely in scope, duration, and conditions, and a central purpose of this survey is to catalog and analyze that variation.

"Derived Data" refers to data, models, scores, indices, analytics outputs, or enriched datasets created by the licensee using the licensor's raw data as an input. This definition is consistent with the definition of "Derived Data" as currently set forth in Section 1.14 of the draft MDLA under negotiation between Quorum and Meridian. The term encompasses a broad range of licensee-created outputs, from simple aggregations of raw data to complex multi-source blended models such as QuorumScore™.

"Revenue Share" refers to a percentage-based payment obligation tied to the licensee's commercialization of Derived Data to third parties. Revenue-share provisions, where present, typically require the licensee to pay the data licensor a specified percentage of net revenues (or, less commonly, gross revenues) generated from the sale, sublicensing, or other commercial distribution of Derived Data products.

These definitions are used throughout this memorandum unless otherwise indicated.

III. Overview of Meridian's Proposed License-Back and Revenue-Share Terms

A. The Three Components

Meridian's proposed license-back and revenue-share package, as set forth in Meridian's redline of Article VII of the draft MDLA (circulated by Graham Whitley of Kessler Pruitt & Ashe LLP on March 8, 2023), consists of three interlocking components:

1. Perpetual, Royalty-Free License-Back. Under proposed Section 7.2(a) of the MDLA, Meridian would receive a perpetual, irrevocable, royalty-free, non-exclusive, worldwide license to use, reproduce, distribute, display, perform, and create derivative works from all Derived Data for Meridian's own internal and commercial purposes. The license-back is not limited to any particular use case, product line, or business unit within Meridian. It expressly survives termination or expiration of the MDLA for any reason, meaning that Meridian would retain access to Quorum's Derived Data even after the parties' commercial relationship ends. The proposed provision contains no restrictions on Meridian's ability to sublicense the Derived Data to its own customers or partners, and no limitations on Meridian's use of the Derived Data in competition with Quorum.

2. Commercialization Gating with Sole Discretion. Under proposed Section 7.3(b), Quorum would be prohibited from selling, sublicensing, distributing, or otherwise making available any Derived Data to any third party unless and until the Derived Data has been "sufficiently transformed" such that "no Meridian Licensed Data can be reverse-engineered, reconstructed, or otherwise derived" from the output. The critical feature of this provision is that Meridian retains "sole and absolute discretion" to determine whether any particular output meets the "sufficient transformation" standard. The draft MDLA does not define "sufficient transformation" by reference to any objective technical criteria, does not provide for independent expert review, and does not include any appeal or dispute-resolution mechanism if Quorum disagrees with Meridian's determination. Quorum would be required to submit each proposed commercial output to Meridian for review at least sixty (60) days prior to any planned distribution, and Meridian would have no obligation to provide reasons for any determination that the output is insufficiently transformed.

3. 15% Revenue Share. Under proposed Section 7.4, if Quorum receives Meridian's written approval to commercialize Derived Data to third parties (i.e., if Meridian determines that the output is "sufficiently transformed"), Meridian would receive 15% of "Net Revenues" attributable to such Derived Data. "Net Revenues" is defined as gross revenues from the commercialization of approved Derived Data, less only direct third-party distribution costs (excluding Quorum's internal development costs, overhead, personnel costs, or marketing expenses). The revenue-share obligation would continue for the term of the MDLA and for a period of three (3) years following termination or expiration. The proposed provision does not include any cap on aggregate revenue-share payments, any sunset or step-down mechanism, or any methodology for attributing revenues to Meridian's data inputs versus Quorum's independent contributions in multi-source blended products.

B. Context — Meridian's Stated Rationale

Meridian's negotiating team has advanced several arguments in support of its license-back and revenue-share demands. As articulated by Arthur Pennington (General Counsel) and Keith Marlow (SVP, Data Licensing & Partnerships), and through Meridian's outside counsel at Kessler Pruitt & Ashe LLP (Graham Whitley), Meridian's position rests on the following premises:

intellectual Property Argument. Meridian contends that Derived Data constitutes a "derivative work" under copyright law (17 U.S.C. § 103) that remains subject to Meridian's underlying intellectual property rights in the raw licensed data. On this view, Quorum cannot create Derived Data without utilizing Meridian's copyrightable compilations and databases, and any output that incorporates or builds upon Meridian's data retains a nexus to Meridian's IP that justifies a license-back. We note that this legal characterization is contested — the copyright status of factual data compilations is limited under Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991), and the "derivative work" analysis for data-driven analytics outputs is unsettled — but a full analysis of the copyright issues is beyond the scope of this market-survey memorandum.

Data Moat and Valuation Narrative. Behind Meridian's formal IP argument lies a commercial and investor-relations motivation. Breckenridge Capital Partners Fund IV, L.P., which holds a 62.5% ownership interest in Meridian (with Caleb Rourke as the partner overseeing the Meridian investment), and Whitmore Zanthor & Co. (with Rachel Ng-Castillo as Managing Director advising Meridian on a potential 2025–2026 initial public offering) have expressed concern that granting broad derived-data rights to licensees like Quorum could undermine Meridian's "data moat" narrative. In the context of a pre-IPO company, the ability to demonstrate that downstream derivative uses of Meridian's data remain within Meridian's commercial orbit — through license-backs, revenue shares, and commercialization controls — supports a higher enterprise valuation by suggesting that Meridian's data generates value not only through direct licensing fees but also through participation in downstream analytics economics. This valuation narrative consideration, while commercially understandable from Meridian's perspective, does not reflect established market norms for data licensing and should not be allowed to distort the contractual terms of the MDLA.

Board Resolution. We note for completeness that Meridian's board of directors formally resolved, on August 22, 2023, against granting unrestricted derived-data commercialization rights in any data-licensing transaction. Although this resolution postdates both the period of our survey and the current stage of negotiations, the commercial dynamic that produced it was evident in earlier negotiations and informs Meridian's posture throughout. The resolution suggests that Meridian's position on license-back terms is unlikely to change absent significant counterpressure from Quorum.

IV. Market Survey Findings — License-Back Provisions

A. Prevalence of License-Back Provisions

Of the approximately thirty-five (35) enterprise data-licensing agreements surveyed, approximately twelve (12) — roughly 34% — contained any form of license-back provision relating to licensee-created derived data. The remaining approximately twenty-three (23) agreements (66%) contained no license-back whatsoever. In those agreements, the licensee retained sole and exclusive ownership of all derived outputs — including models, scores, indices, enriched datasets, and blended analytics products — subject only to standard restrictions on reverse engineering of the licensor's raw data, re-identification of de-identified data, and redistribution of the licensor's raw or unmodified data.

This finding is significant. It establishes that license-back provisions are a minority practice in enterprise data licensing, not a market standard. The majority of data licensors in the surveyed population treated their compensation as fully captured by the base license fee and did not seek additional rights in licensee-created derived outputs. In the majority-practice agreements, the parties typically addressed the derived-data issue through a combination of: (i) a clear ownership clause vesting all IP in derived outputs in the licensee; (ii) restrictions on redistribution of the licensor's raw data in unmodified or minimally modified form; (iii) prohibitions on reverse engineering of the licensor's data from derived outputs; and (iv) in some cases, non-compete or non-solicitation provisions restricting the licensee from using the licensor's data to build competing data products that would directly compete with the licensor's core data offerings. This majority-practice framework draws a clear distinction between the licensor's proprietary rights in its raw data (which are protected) and the licensee's proprietary rights in its derived outputs (which are respected), without requiring any license-back mechanism.

Conclusion: The existence of any license-back provision in the Quorum-Meridian MDLA would place the agreement in the minority of observed transactions. Meridian's proposal, which demands a broad and commercially unrestricted license-back, is further outside the mainstream.

B. Scope and Character of License-Back Provisions Where Present

Among the twelve (12) agreements that contained some form of license-back provision, the scope and character of the license-back varied substantially. We observed two broad categories:

1. Narrow, Internal-Use-Only License-Backs (8 of 12 Agreements)

The predominant form of license-back — found in eight (8) of the twelve (12) agreements containing any license-back — was a narrow, internal-use-only license limited in scope, purpose, and duration. These provisions shared the following common characteristics:

• Limited Subject Matter. The licensor received a non-exclusive, non-transferable license to use only aggregated or anonymized derived outputs, such as aggregate statistical models, benchmarking indices, or de-identified performance metrics. The license-back did not extend to the licensee's specific derived data products, proprietary models, algorithms, scoring methodologies, or enriched datasets. In several agreements, the license-back was expressly limited to "aggregate statistical outputs that do not identify or permit identification of any individual data subject or any specific output of [licensee's] proprietary models."

• Internal-Use-Only Restriction. The licensor was permitted to use the licensed-back outputs solely for its own internal purposes, typically limited to product improvement, research and development, benchmarking, and quality assurance. The licensor was expressly prohibited from using the licensed-back outputs for commercial purposes, including incorporating them into the licensor's own products or services offered to third parties, distributing them to the licensor's other customers, or using them for competitive purposes.

• Limited Duration. In most cases, the license-back was not perpetual. Six (6) of the eight (8) narrow license-backs were coterminous with the term of the underlying data license agreement, terminating upon expiration or termination of the main agreement. The remaining two (2) included defined survival periods of two to three years post-termination, after which the licensor's right to use the licensed-back outputs expired.

• No Sublicensing Right. None of the eight (8) narrow license-backs granted the licensor any right to sublicense the licensed-back outputs to third parties.

• No Revenue-Share Linkage. None of these eight agreements also included a revenue-share provision. The license-back was treated as a self-contained mechanism, and the licensor's compensation was limited to the base license fee plus the limited license-back right.

As a representative example, one credit-data licensing arrangement in the survey involved a major credit bureau licensing raw credit data to a financial services company that used the data to build proprietary risk-scoring models. The license-back provision in that agreement granted the credit bureau a non-exclusive, non-transferable, internal-use-only license limited to "anonymized aggregate model performance metrics" — i.e., high-level statistical information about how the licensee's models performed (e.g., overall accuracy rates, distribution of scores across population segments) — for the sole purpose of "improving and enhancing [licensor's] own scoring products and methodologies." The license-back did not extend to the licensee's proprietary models, algorithms, individual scores, or enriched datasets. It was coterminous with the license term and contained no sublicensing right.

2. Broader License-Backs (4 of 12 Agreements)

Four (4) of the twelve (12) agreements containing license-back provisions included broader license-back grants that permitted the licensor to use licensee-created derived outputs for commercial purposes (not limited to internal use). However, even in these agreements — which represent the outer boundary of observed market practice — the license-back was subject to significant limitations, conditions, and reciprocal compensation that distinguish them sharply from Meridian's proposal:

• Negotiated Carve-Outs. In all four agreements, the license-back was subject to specific, negotiated carve-outs that excluded the licensee's most commercially sensitive derived outputs from the scope of the license-back. Common carve-outs included: (i) the licensee's proprietary models, algorithms, and trade secrets; (ii) outputs that incorporated data from sources other than the licensor (i.e., multi-source blended outputs); and (iii) outputs that the licensee designated as "competitively sensitive" pursuant to agreed criteria. These carve-outs ensured that the licensor's license-back right did not extend to the licensee's core proprietary analytics products.

• Reciprocal Compensation. In three of the four agreements, the broader license-back was not royalty-free. Instead, it was accompanied by reciprocal compensation flowing from the licensor to the licensee, structured as fee reductions (Discounts on the base license fee), expanded data access (the licensee received access to additional data sets or higher-frequency data feeds at no additional cost), or co-branding credits (the licensor agreed to co-brand certain products utilizing the licensed-back data, with the licensee receiving attribution and co-marketing benefits). The fourth agreement provided that the licensor would pay a separately negotiated fee for any commercial use of the licensed-back outputs, determined on a case-by-case basis. In no case was the broader license-back granted on a purely royalty-free basis.

• Joint or Expert Determination. In all four agreements, the scope of the license-back — including what constituted "derived data" subject to the license-back and what fell within the carve-outs — was determined through a jointly agreed methodology or, in the event of disagreement, by an independent expert. No agreement granted the licensor sole discretion to determine the scope of the license-back or the eligibility of particular outputs.

• Duration Limitations. Two of the four broader license-backs were coterminous with the license term. The other two included defined survival periods (one year and three years, respectively) post-termination. None was perpetual with respect to specific models, scores, or enriched datasets.

Critical Finding: No surveyed agreement granted the licensor a perpetual, royalty-free, unrestricted commercial license to all licensee-created derived data without carve-outs, reciprocal compensation, or scope limitations determined by objective criteria or independent review. Meridian's proposed license-back is broader in scope, longer in duration, and freer of conditions than any license-back provision observed in the survey.

C. License-Back Duration

Duration is a critical dimension of license-back provisions, as it determines whether the licensor's rights in the licensee's derived outputs extend beyond the parties' commercial relationship. Among the twelve (12) agreements containing license-back provisions, the duration breakdown was as follows:

Perpetual license-back provisions were rare in the surveyed population. In the two agreements where a perpetual license-back element was present, the perpetual right was strictly limited to fully anonymized aggregate statistical data — e.g., aggregate benchmarking indices or high-level population-level statistical outputs — that could not be attributed to any individual data subject or used to reconstruct the licensee's proprietary models or specific derived products. These perpetual elements were understood by the parties as granting the licensor a right to retain high-level market intelligence derived from the licensing relationship, not as granting the licensor ongoing access to the licensee's proprietary analytics outputs.

Conclusion: Meridian's demand for a perpetual license-back covering all Derived Data — including proprietary models, scores, enriched datasets, and multi-source blended outputs like QuorumScore™ — is outside the observed range of market practice. Even the small number of agreements with any perpetual license-back element limited that element to anonymized aggregates, a far cry from Meridian's proposal.

D. Licensor Discretion Over Commercialization

Meridian's proposal to retain "sole and absolute discretion" to determine whether Quorum's derived outputs are "sufficiently transformed" for third-party commercialization is, in our assessment, the most anomalous element of the license-back package. This section reports our findings on commercialization-gating mechanisms in the surveyed agreements.

No surveyed agreement granted the data licensor sole discretion to determine whether licensee-created outputs were "sufficiently transformed" for third-party commercialization. Where transformation, de-identification, or independence-from-raw-data standards were included as a condition of the licensee's right to commercialize derived outputs — which occurred in approximately eight (8) of the thirty-five (35) agreements — the determination was made through one of the following mechanisms:

(i) Objective, measurable technical criteria specified in the agreement (5 agreements). These agreements defined "sufficient transformation" or its equivalent by reference to specific, quantifiable standards — for example, requiring that the derived output not permit reconstruction of any individual data point from the licensor's raw data with a specified confidence level, or requiring that the derived output incorporate data from at least a specified number of independent sources constituting at least a specified percentage of total data inputs by volume. These criteria were set forth in the agreement (often in a technical schedule or exhibit) and were testable by either party without requiring the licensor's subjective judgment.

(ii) A jointly appointed independent technical expert (2 agreements). These agreements provided that any dispute over whether a derived output met the applicable transformation standard would be resolved by an independent technical expert jointly selected by the parties (or, if the parties could not agree on an expert, appointed by a designated neutral institution). The expert's determination was binding on both parties. This mechanism ensured that neither party could unilaterally block the other's commercialization efforts.

(iii) Mutual agreement of the parties, with escalation to an independent expert (1 agreement). This agreement required the parties to attempt to reach mutual agreement on whether the derived output met the applicable standard, with escalation to an independent expert if mutual agreement could not be reached within a specified period (in this case, thirty (30) days).

In none of the surveyed agreements did the licensor possess a unilateral veto over the licensee's commercialization of derived outputs. The absence of sole-discretion gatekeeping in the surveyed population is not surprising. Industry commentators and practitioners have consistently identified sole-discretion standards as commercially impracticable and potentially abusive in the data-licensing context. Where the data licensor also operates in downstream analytics or scoring markets — as Meridian does — sole discretion to approve or reject the licensee's commercial outputs creates an inherent conflict of interest: the licensor has an incentive to withhold approval in order to prevent the licensee from competing in downstream markets, regardless of whether the output is genuinely "insufficiently transformed." This dynamic is sometimes referred to as "hold-up risk" and is widely recognized as a significant commercial concern in data-licensing negotiations.

The practical effect of Meridian's sole-discretion proposal would be to give Meridian an effective veto over Quorum's downstream commercial strategy. Because "sufficient transformation" is undefined in the MDLA and Meridian would have no obligation to provide reasons for a negative determination, Quorum would have no ability to predict whether any particular output would be approved, no ability to challenge a negative determination, and no recourse in the event that Meridian exercises its discretion in a manner designed to protect its own competitive position rather than to address legitimate concerns about the integrity of its raw data. This risk is particularly acute given the competitive overlap between Meridian and Quorum in downstream analytics markets.

Conclusion: Sole-discretion commercialization gatekeeping by the data licensor has no precedent in the surveyed agreements and is inconsistent with market practice. The observed market mechanisms for addressing transformation and de-identification standards uniformly rely on objective criteria, independent expert review, or mutual agreement — not unilateral licensor determination.

V. Market Survey Findings — Revenue-Sharing Arrangements

A. Prevalence

Revenue-sharing provisions tied to the licensee's commercialization of derived data were a minority feature in the surveyed agreements. Approximately six (6) of the thirty-five (35) agreements surveyed — roughly 17% — included any form of revenue-sharing mechanism. The remaining twenty-nine (29) agreements (83%) allocated all commercial value from derived data entirely to the licensee, with the data licensor's compensation limited to the base license fee (whether structured as a flat annual fee, a per-record fee, a tiered usage fee, or a combination thereof).

This finding reinforces the conclusion reached in Part IV.A above regarding the minority status of license-back provisions. The majority of data licensors in the surveyed population did not seek to participate in the economics of their licensees' derived-data commercialization beyond the base license fee. The base license fee was understood by the parties as full compensation for the licensor's contribution of raw data, and the licensee's investment in developing derived analytics products — including the cost of proprietary data inputs, model development, algorithm design, testing, validation, and commercialization — was treated as the licensee's investment yielding the licensee's returns.

B. Revenue-Share Rates Where Present

Among the six (6) agreements that included revenue-sharing provisions, the observed rates were as follows:

The median observed revenue-share rate was approximately 5–7% of net revenues. The range extended from 3% at the low end to 12% at the high end. The highest observed rate of 12% applied in a single agreement where the licensor's raw data constituted the predominant input to the derived product — specifically, where the licensor's data represented more than 80% of total data inputs by volume and the licensee's independent contribution was limited to aggregation, formatting, and basic statistical analysis without substantial proprietary modeling or blending with independent data sources. This fact pattern is materially different from the Quorum-Meridian arrangement, in which Quorum blends Meridian credit data with its own extensive proprietary first-party claims, policy, and telematics data and applies proprietary machine learning models and algorithms to produce QuorumScore™ outputs.

No surveyed agreement imposed a revenue-share rate of 15% or higher. Meridian's proposed 15% rate exceeds the top of the observed range by a margin of three percentage points and is approximately double the median rate. Even if one were to accept that some revenue share is appropriate in the Quorum-Meridian context (a proposition that we do not concede), the 15% rate demanded by Meridian is not supported by market practice.

C. Structure of Revenue-Share Provisions

Beyond the headline rate, the structural features of revenue-share provisions in the surveyed agreements are relevant to assessing Meridian's proposal. In all six (6) agreements containing revenue-share provisions, the revenue share was subject to conditions and limitations that are absent from Meridian's proposed terms:

• Caps. Five of the six agreements included caps on aggregate revenue-share payments, structured either as absolute dollar amounts (e.g., total revenue-share payments not to exceed a specified dollar amount per year or over the term of the agreement) or as a percentage of the total license fee (e.g., aggregate revenue-share payments not to exceed a specified percentage of the cumulative base license fee paid by the licensee). These caps ensured that the revenue-share obligation remained proportionate to the overall economics of the licensing relationship and did not grow to dominate the licensee's cost structure.

• Sunset Provisions. Four of the six agreements included sunset provisions under which the revenue-share rate declined over time or terminated altogether after a defined period, typically three to five years from the commencement of the licensee's commercialization of derived data. The rationale for sunset provisions is straightforward: over time, the licensee's independent investment in developing and enhancing the derived product increasingly dominates the licensor's initial data contribution, and the revenue share should diminish accordingly.

• Attribution Methodology. All six agreements included a clear, agreed-upon methodology for attributing revenues to derived data incorporating the licensor's inputs versus revenues attributable to the licensee's independent contributions. In multi-source blended products — such as a product that combines data from multiple licensors with the licensee's own proprietary data — the attribution methodology determined what portion of total revenues was subject to the revenue share. Common approaches included: (a) allocation based on relative data-input volumes; (b) allocation based on an agreed model-contribution analysis; and (c) allocation based on a jointly conducted attribution study using statistical techniques to isolate the marginal contribution of each data input.

• Good-Faith-Efforts Obligations. The revenue-share provisions were structured as obligations to pay a percentage of actual revenues, not as guaranteed minimum payments. The licensee was not obligated to achieve any particular level of commercialization revenues and was subject only to good-faith-efforts obligations to pursue commercialization opportunities.

Two additional structural features merit emphasis:

First, in all six (6) agreements with revenue-share provisions, the licensee retained sole ownership of the intellectual property in the derived data. The revenue share was characterized as additional consideration for the underlying data license — i.e., an incremental fee reflecting the licensor's contribution to the licensee's downstream commercial success — not as evidence of licensor co-ownership of, or co-authorship interest in, the derived outputs. This distinction is important because it confirms that a revenue share, even where present, does not imply or create any ownership interest by the licensor in the licensee's derived products.

Second, no surveyed agreement combined a perpetual royalty-free license-back with a revenue share. These two mechanisms serve overlapping commercial purposes: each is a means by which the licensor participates in the value created through licensee-derived outputs. The license-back gives the licensor direct access to use the derived outputs; the revenue share gives the licensor a financial stake in the licensee's commercialization of the derived outputs. Combining both mechanisms would effectively double-count the licensor's claimed interest, permitting the licensor to both use the derived outputs for its own commercial benefit (through the license-back) and extract a financial toll on the licensee's use of the same outputs (through the revenue share). This combination was not observed in any of the surveyed agreements, and its commercial logic is difficult to justify.

D. QuorumScore™ Implications

The implications of Meridian's revenue-share proposal for Quorum's planned QuorumScore™ B2B licensing business merit specific attention.

If Meridian's proposed 15% revenue-share rate were applied to Quorum's planned third-party licensing of QuorumScore™ outputs, the financial impact would be material. At the projected revenue level of $40–50 million in annual B2B licensing revenues by 2027, the revenue share would amount to approximately $6.0–7.5 million annually payable to Meridian. This sum is substantial both in absolute terms and relative to the overall economics of the Quorum-Meridian licensing relationship. For comparison, the base license fee under the proposed MDLA is $60 million over the five-year term ($10.8 million in Year 1 and $12.3 million in each of Years 2 through 5). A $6–7.5 million annual revenue-share payment would represent approximately 50–61% of the annual base license fee — effectively transforming Meridian from a data supplier into a de facto revenue partner in Quorum's downstream analytics business, notwithstanding that Quorum bears all of the development cost, commercialization risk, and competitive effort associated with building and marketing QuorumScore™.

Under the surveyed market practice, a blended model like QuorumScore™ — which combines Meridian credit data with Quorum's proprietary first-party claims, policy, and telematics data, and applies Quorum's proprietary machine learning models and algorithms — would typically be treated as predominantly a licensee creation. The data licensor's contribution of raw credit data would be addressed through the base license fee (here, $60 million over five years), and the licensee would retain all commercial rights in the blended output without further payment obligation. Even in the minority of agreements that included revenue-share provisions, carve-outs for the licensee's proprietary models and multi-source blended outputs were common. Several agreements expressly excluded from the revenue-share obligation any product in which the licensor's data constituted less than a specified percentage (ranging from 30% to 50%) of total data inputs, reflecting the view that where the licensee's independent contribution is substantial, the base license fee adequately compensates the data supplier.

The application of a 15% revenue share to QuorumScore™ — a product in which Quorum's independent first-party data and proprietary modeling constitute a significant (and arguably predominant) share of the total input value — is inconsistent with the structural logic observed in the surveyed agreements and would impose a disproportionate cost on Quorum relative to Meridian's actual data contribution.

VI. Market Survey Findings — Post-Termination Treatment of Derived Data

A. Observed Approaches

The treatment of licensee-created Derived Data upon termination or expiration of the data-licensing agreement is a critical commercial term and is currently a disputed point in the Quorum-Meridian MDLA negotiation. Our survey identified three approaches, with varying frequency:

Most Common Approach — Licensee Retains Derived Data (approximately 20 of 35 agreements). The most common approach, observed in approximately twenty (20) of the thirty-five (35) surveyed agreements, permitted the licensee to retain ownership of and continue to use Derived Data following termination or expiration of the agreement, subject to two conditions: (i) all raw licensor data must be returned to the licensor or destroyed within a specified period (commonly thirty (30) to ninety (90) days) following the effective date of termination, with the licensee providing a certified officer's certificate confirming return or destruction; and (ii) the Derived Data retained by the licensee must not contain or permit reconstruction of the licensor's raw data. This approach reflects the principle that the licensee's investment in creating Derived Data — including development costs, proprietary model design, and integration of independent data sources — generates an asset that belongs to the licensee and should not be forfeited upon termination of the data supply relationship. The licensor's interest in protecting its raw data is addressed through the return/destruction obligation, not through forfeiture of the licensee's derived outputs.

Less Common Approach — Continued Use with Wind-Down (approximately 10 of 35 agreements). In approximately ten (10) of the surveyed agreements, the licensee was permitted to retain Derived Data following termination, but subject to continued compliance with use restrictions and confidentiality obligations from the agreement, together with a defined wind-down period (typically six (6) to twelve (12) months) for transitioning away from commercial use of the Derived Data. During the wind-down period, the licensee could continue to use Derived Data for existing commercial commitments (e.g., fulfilling outstanding customer contracts) but could not enter into new commercial arrangements involving the Derived Data. At the end of the wind-down period, the licensee's right to use the Derived Data for commercial purposes terminated, although the licensee was typically permitted to retain the Derived Data for archival, regulatory compliance, and audit purposes.

Rare Approach — Destruction of Derived Data (approximately 5 of 35 agreements). In approximately five (5) of the surveyed agreements, the licensee was required to destroy Derived Data along with raw licensed data upon termination. This approach was observed primarily in highly regulated contexts — most notably, agreements governed by HIPAA requirements for protected health information and agreements involving classified government data — and was not observed in any credit-data or financial-data licensing arrangement comparable to the Quorum-Meridian transaction. In these agreements, the destruction requirement was driven by regulatory mandates rather than by the parties' commercial preferences.

Relevance to the Quorum-Meridian Negotiation. The MDLA currently treats the post-termination treatment of Derived Data as a disputed term. Meridian's position is that all Derived Data must be returned to Meridian or destroyed upon termination, treating Derived Data the same as raw licensed data for purposes of the post-termination return/destruction obligation. Quorum's position is that it should retain the right to continue using Derived Data post-termination, subject to return or destruction of all raw Meridian data within ninety (90) days. The market survey findings support Quorum's position: the majority approach (approximately 57% of surveyed agreements) permits the licensee to retain and continue using Derived Data post-termination, and the destruction-of-derived-data approach was limited to regulatory contexts not applicable here.

B. Interaction with License-Back

The interaction between license-back provisions and post-termination treatment of Derived Data raises a distinct concern specific to Meridian's proposal. In the small number of surveyed agreements that contained both a license-back provision and provisions addressing post-termination treatment of Derived Data, the license-back typically terminated upon expiration of the main agreement, consistent with the general pattern of coterminous license-back provisions described in Part IV.C above. The limited exception was for license-backs covering fully anonymized aggregate statistical data, which in two cases survived termination on a perpetual basis (as noted above).

Meridian's proposal creates a significant asymmetry. If the perpetual license-back demanded by Meridian is combined with Meridian's separate demand that Quorum destroy its Derived Data upon termination, the result would be a one-sided outcome in which:

(a) Meridian retains perpetual access to all Derived Data created by Quorum during the term of the MDLA, including QuorumScore™ outputs, for Meridian's own internal and commercial purposes; while

(b) Quorum loses access to its own derived creations — the products of Quorum's substantial investment in proprietary data, model development, and analytics capabilities — and is required to destroy them.

This outcome would effectively transfer the value of Quorum's derived-data portfolio to Meridian at no cost. It would mean that Meridian, following termination, could continue to use and commercialize outputs that Quorum developed using Quorum's own proprietary inputs and expertise, while Quorum itself would be prohibited from doing so. This asymmetry was not observed in any surveyed agreement and is, in our assessment, commercially unreasonable.

VII. Analytical Framework — Factors Driving License-Back Terms

The market survey findings reveal that the presence, scope, and terms of license-back provisions in enterprise data-licensing agreements are influenced by several identifiable factors. Understanding these factors provides context for assessing the reasonableness of Meridian's demands and for positioning Quorum's negotiating arguments.

A. Relative Contribution of Inputs

The most significant factor influencing the prevalence and scope of license-back provisions is the relative contribution of the licensor's data versus the licensee's independent inputs to the creation of derived outputs.

Agreements in which the licensor's data constituted the dominant input — by volume, by analytical significance, or both — to the derived outputs were more likely to include license-back provisions, and those provisions tended to be broader in scope. Conversely, where the licensee combined the licensor's data with substantial independent proprietary data, models, or algorithms, license-back provisions were less common and, where present, more narrowly scoped. This pattern reflects a commonsense intuition: the greater the licensee's independent contribution, the less justified is the licensor's claim to rights in the resulting output.

Application to Quorum-Meridian. QuorumScore™ is a multi-source blended model that combines Meridian credit data with Quorum's own proprietary first-party data sets — including historical claims data accumulated over more than two decades, policy-level underwriting data, and telematics data collected from Quorum's policyholder base — as well as Quorum's proprietary machine learning models and algorithms developed by Quorum's internal data science team. Quorum's independent contribution to QuorumScore™ is substantial by any measure. This fact pattern favors a narrower (or no) license-back under the analytical framework observed in the market, and it undercuts Meridian's assertion that its credit data contribution justifies a broad, unrestricted license-back on all Derived Data.

B. Competitive Dynamics

The competitive relationship between the licensor and the licensee is a second important factor. License-back provisions were more commonly observed in agreements between parties that operated in different market segments with limited competitive overlap. In those settings, granting the licensor a license-back to derived outputs posed less risk to the licensee's competitive position, because the licensor was unlikely to use the derived outputs to compete directly with the licensee's products or services.

By contrast, where the licensor competed with the licensee in downstream analytics or scoring markets, license-back provisions were rare and, where present, were subject to strict use restrictions and competitive-firewall obligations designed to prevent the licensor from using the licensee's derived outputs to gain a competitive advantage. Such restrictions included: prohibitions on the licensor's use of licensed-back data in any product or service that competed with the licensee's offerings; organizational firewalls within the licensor requiring that licensed-back data be accessible only to designated personnel who did not work on competing products; and audit rights allowing the licensee to verify compliance with competitive-use restrictions.

Application to Quorum-Meridian. Meridian operates in the credit-data analytics and scoring market, offering its own credit-scoring products and analytics services to financial institutions, insurers, and other enterprises. Quorum's planned B2B commercialization of QuorumScore™ and related analytics products will position Quorum as a competitor to Meridian in certain downstream analytics markets, particularly in the insurance and underwriting analytics space. This competitive overlap counsels strongly against granting Meridian a broad license-back, and it renders the sole-discretion commercialization-gating mechanism especially problematic. A provision that gives a competitor sole discretion to approve or reject Quorum's commercial outputs is not merely unusual — it is a structural competitive advantage that no rational licensee would grant absent extraordinary countervailing consideration.

C. IPO and Sponsor Dynamics

A third factor — less frequently discussed in industry commentary but observable in the surveyed data — relates to the ownership structure and capital-markets trajectory of the data licensor. In situations where a data licensor was backed by a private equity sponsor and preparing for an initial public offering, license-back demands were sometimes driven by the licensor's desire to demonstrate to prospective public investors that its data "moat" extends to downstream derivative uses. The ability to point to license-back provisions, revenue-share arrangements, and commercialization controls in the licensor's customer agreements supports a narrative that the licensor's data generates value not only through direct licensing fees but also through participation in the broader ecosystem of derived analytics. This narrative, in turn, supports a higher enterprise valuation and a more favorable IPO pricing.

While this motivation is commercially rational from the licensor's perspective — and we do not suggest that Meridian or its advisors are acting improperly in pursuing it — it is important to recognize that these demands are driven by the licensor's valuation narrative, not by established market norms for data-licensing terms. The cost of supporting the licensor's IPO story should not be borne by the licensee in the form of contractual terms that exceed market practice and impair the licensee's commercial position.

Application to Quorum-Meridian. This dynamic appears to be directly at play in the Quorum-Meridian negotiation. Breckenridge Capital Partners Fund IV, L.P. holds a 62.5% ownership interest in Meridian, and Caleb Rourke (the Breckenridge partner overseeing the Meridian investment) has been actively involved in the commercial terms of the MDLA negotiation. Whitmore Zanthor & Co., with Rachel Ng-Castillo as Managing Director, is advising Meridian on a potential 2025–2026 IPO. The connection between Meridian's aggressive license-back demands and the IPO valuation narrative is, in our assessment, not coincidental. Quorum should be prepared to identify this dynamic in negotiations — not as an accusation, but as a basis for distinguishing Meridian's demands from genuine market practice and for framing Quorum's counterproposal as the commercially normative position.

VIII. Comparative Analysis — Meridian's Proposal vs. Market Practice

A. Summary Comparison Table

The following table summarizes the comparison between Meridian's proposed license-back and revenue-share terms and the market practice observed in the survey:

B. Narrative Assessment

The comparative analysis confirms that Meridian's proposed license-back package is materially outside market norms, both in each individual element and in the aggregate combination of elements.

Individual Element Analysis. Each component of Meridian's proposal, considered in isolation, exceeds the range of market practice observed in the survey:

• The existence of any license-back places the agreement in the minority of observed transactions (34%).

• The scope of the license-back — perpetual, royalty-free, unrestricted as to use, and covering all Derived Data without carve-outs — exceeds even the broadest license-backs observed in the survey, all of which were subject to carve-outs, use restrictions, or reciprocal compensation.

• The duration — perpetual, surviving termination — exceeds all but two observed license-backs, both of which applied their perpetual element only to fully anonymized aggregate statistics.

• The sole-discretion commercialization-gating mechanism has no precedent in any surveyed agreement. Every observed transformation or de-identification standard was determined by objective criteria, independent expert review, or mutual agreement — never by the sole discretion of the data licensor.

• The 15% revenue-share rate exceeds the top of the observed range (12%) and is approximately double the median rate (5–7%). Even the 12% high-water mark applied in a fact pattern (licensor data constituting >80% of inputs) that is materially distinguishable from the Quorum-Meridian arrangement.

Combination Analysis. Beyond the individual elements, the combination of these provisions is itself unprecedented. No surveyed agreement combined a perpetual royalty-free license-back with a revenue share. These mechanisms serve overlapping purposes, and their combination effectively imposes a double cost on the licensee: the licensor obtains both direct access to the derived outputs (through the license-back) and a financial toll on the licensee's commercialization of the same outputs (through the revenue share). When the sole-discretion gatekeeping mechanism is added, the combination becomes triply burdensome: Meridian would receive free access to Quorum's derived outputs, a 15% share of Quorum's commercialization revenues, and an unreviewable veto over whether Quorum may commercialize at all.

Most Anomalous Element. The sole-discretion commercialization-gating mechanism is, in our assessment, the most commercially problematic element of Meridian's proposal. While a license-back can be negotiated to a narrow and tolerable scope, and a revenue-share rate can be negotiated downward, the sole-discretion gatekeeping mechanism poses a structural risk that cannot be mitigated through rate adjustments: it gives Meridian an effective veto over Quorum's downstream commercial strategy, exercisable without constraint, without transparency, and without recourse. This mechanism transforms the licensing relationship from a data-supply arrangement into a quasi-joint-venture in which one party holds unchecked authority over the other's commercial activities — a dynamic that is fundamentally inconsistent with the arm's-length data-licensing model and that has no support in observed market practice.

IX. Recommendations

Based on the market survey findings and comparative analysis set forth above, we recommend the following negotiating positions for Quorum, structured as a primary position and two fallback positions in order of preference.

A. Primary Position: No License-Back, No Revenue Share

Consistent with the majority market practice observed in our survey — in which 66% of agreements contain no license-back and 83% contain no revenue share — Quorum should continue to assert that it owns all intellectual property in Derived Data, with unrestricted rights to commercialize Derived Data to third parties, and that no license-back or revenue share to Meridian is required.

This position rests on the following grounds:

1. Adequate Compensation Through Base License Fee. Meridian is adequately compensated for its contribution of raw credit data through the base license fee of $60,000,000 over the five-year term of the MDLA ($10.8 million in Year 1 and $12.3 million in each of Years 2 through 5). The base license fee reflects the market value of Meridian's data and represents full consideration for all rights granted to Quorum under the MDLA, including the right to use Meridian's data as an input to the creation of Derived Data. The Pinnacle Ridge Advisory LLC pricing benchmark study (delivered February 14, 2023, prepared by Franklin Osei) contains comparative data on license economics that supports this position.

2. Quorum's Substantial Independent Contribution. Quorum contributes substantial independent proprietary data (claims, policy, and telematics data accumulated over more than two decades), proprietary machine learning models and algorithms developed by Quorum's internal data science team, and significant development resources to the creation of Derived Data products, including QuorumScore™. This independent contribution is the primary source of value in Quorum's derived analytics products, and Quorum should own and control that value without encumbrance.

3. Market-Practice Support. The primary position is the majority-practice position in the surveyed market, adopted by approximately two-thirds of comparable agreements. It is the presumptive norm in enterprise data licensing, not an aggressive or outlier position.

4. Protection of QuorumScore™. The primary position protects Quorum's ability to develop, commercialize, and license QuorumScore™ outputs without restriction, preserving the projected $40–50 million B2B revenue line and the value of Quorum's pending trademark application (USPTO Serial No. 97/482,116).

B. Fallback Position: Narrow, Time-Limited License-Back Without Revenue Share

If a license-back is commercially necessary to close the MDLA transaction — and we recognize that Meridian has indicated through counsel that some form of license-back may be a condition to closing — Quorum should negotiate for a license-back provision that falls within the range of observed market practice, specifically:

(a) Scope — In

PRIVILEGED AND CONFIDENTIAL — ATTORNEY-CLIENT COMMUNICATION / ATTORNEY WORK PRODUCT This memorandum is protected by the attorney-client privilege and the work product doctrine. It is intended solely for the use of the addressees identified herein in connection with Calderwood & Harkness LLP's representation of Quorum Insurance Holdings Corp. Any disclosure, distribution, or copying of this memorandum without the express written consent of Calderwood & Harkness LLP is strictly prohibited.
MEMORANDUM
TO: Wendy Schultz, Associate, intellectual Property, Calderwood & Harkness LLP Leonard Cross, Partner, intellectual Property, Calderwood & Harkness LLP
FROM: Omar Siddiqui, Associate, intellectual Property, Calderwood & Harkness LLP
DATE: April 2023
RE: Market Survey — License-Back Provisions in Enterprise Data-Licensing Transactions
MATTER: Quorum Insurance Holdings Corp. — Meridian Credit intelligence, Inc. Data License Negotiation; C&H Matter No. 1025-00008
CC: Patricia Heng, Senior Counsel, Technology Transactions, Calderwood & Harkness LLP
Right-click to update Table of Contents
I. Executive Summary
This memorandum presents the results of a market survey of license-back provisions in enterprise data-licensing transactions, prepared in connection with Calderwood & Harkness LLP's representation of Quorum Insurance Holdings Corp. ("Quorum") in its ongoing negotiation of the proposed Master Data License Agreement ("MDLA") with Meridian Credit intelligence, Inc. ("Meridian"). The survey focuses specifically on arrangements in which a licensee creates derived data — including blended risk scores, propensity models, enriched datasets, and other derivative analytics outputs — from licensed third-party data, and the data licensor claims rights to such derived outputs through license-back provisions, commercialization restrictions, or revenue-sharing mechanisms.
The impetus for this survey is Meridian's demand, communicated through its outside counsel at Kessler Pruitt & Ashe LLP (Graham Whitley) and its internal negotiating team led by Arthur Pennington (General Counsel) and Keith Marlow (SVP, Data Licensing & Partnerships), for a license-back and revenue-sharing package relating to Derived Data (as that term is defined in the current draft of the MDLA). Meridian's proposed terms consist of three components:
(a) Perpetual, Royalty-Free License-Back. Meridian would receive a perpetual, royalty-free, non-exclusive license-back on all Derived Data created by Quorum using Meridian's licensed data as an input. This license-back would permit Meridian to use all such Derived Data for its own internal and commercial purposes and would survive termination of the MDLA.
(b) Commercialization Gating with Sole Discretion. Quorum would be prohibited from selling or sublicensing Derived Data to third parties unless the Derived Data has been "sufficiently transformed" such that no Meridian raw data can be reverse-engineered from the output. Critically, Meridian would retain sole discretion to determine whether any particular output meets the "sufficient transformation" standard.
(c) 15% Revenue Share. If Quorum receives Meridian's approval to commercialize Derived Data to third parties, Meridian would receive 15% of net revenues attributable to such Derived Data.
Overall Conclusion. Based on the market survey detailed below, Meridian's proposed license-back package — combining a perpetual royalty-free license-back, sole-discretion transformation gatekeeping, and a 15% revenue share — is materially outside market norms for enterprise data-licensing transactions, both when each element is considered individually and, most strikingly, when the elements are considered in combination. No surveyed agreement combined all three of these mechanisms, and several individual components of Meridian's proposal (particularly the sole-discretion gatekeeping mechanism and the 15% revenue-share rate) have no precedent in the agreements we reviewed.
Specific Strategic Concern. The license-back proposal poses a direct threat to Quorum's strategic investment in QuorumScore™, a proprietary predictive underwriting model (USPTO trademark application pending, Serial No. 97/482,116) that blends Meridian credit data with Quorum's first-party claims, policy, and telematics data. QuorumScore™ is central to Quorum's planned B2B licensing revenue line, which is projected to generate $40–50 million in annual revenues by 2027. Under Meridian's proposed terms, Quorum's ability to commercialize QuorumScore™ outputs to third parties would be subject to Meridian's sole-discretion approval, and even if approved, would be burdened by a 15% revenue share that would transfer $6–7.5 million annually to Meridian — a material sum that would fundamentally alter the economics of the analytics strategy. Moreover, Meridian would receive a perpetual, royalty-free license to use QuorumScore™ outputs for its own commercial purposes, including in markets where Meridian competes directly with Quorum's planned downstream analytics offerings.
This memorandum sets forth the detailed findings of the market survey, provides a comparative analysis of Meridian's proposal against observed market practice, and offers recommended negotiating positions for Quorum.
II. Scope and Methodology
A. Scope of Survey
This survey examined approximately thirty-five (35) enterprise data-licensing agreements spanning the credit-data, financial-data, consumer-data, and analytics sectors. The agreements were sourced from a combination of publicly available materials and proprietary research, including:
• Agreements filed as exhibits to SEC registration statements, annual reports, and current reports (Forms S-1, 10-K, 10-Q, and 8-K), as well as agreements produced in connection with regulatory submissions to state insurance departments and federal banking regulators;
• Proprietary databases maintained by Calderwood & Harkness LLP's Technology Transactions and intellectual Property practice groups, including transaction databases reflecting the firm's representation of clients in comparable data-licensing transactions;
• Reported litigation involving disputes over derived-data ownership, license-back provisions, and revenue-sharing obligations in data-licensing arrangements, including court opinions and publicly filed pleadings that contained or described relevant contractual terms; and
• Industry commentary and guidance materials published by trade associations (including the Data & Marketing Association and the International Association of Privacy Professionals), law firm client memoranda, and academic literature addressing market practice in data licensing.
The survey focused specifically on agreements in which the licensee uses licensed data as an input to create derivative analytics products, models, scores, or enriched datasets — i.e., transactions in which the derived-data ownership question is commercially material. We excluded from the survey two categories of agreements that present different commercial dynamics and are not directly comparable to the Quorum-Meridian transaction:
• Pure reseller/redistribution licenses, in which the licensee does not create derived outputs but instead redistributes the licensor's data (or lightly formatted versions thereof) to end users. In such arrangements, the licensor typically retains ownership of all data and the licensee acts as a distribution channel, with economics structured as a resale margin rather than a license fee. These arrangements do not raise the derived-data ownership issues central to the Quorum-Meridian negotiation.
• Intra-group data-sharing arrangements, in which data flows between affiliated entities within a single corporate group. These arrangements are governed by different considerations (including tax, regulatory, and intercompany pricing requirements) and do not reflect arm's-length market terms.
The time period covered by the survey encompasses agreements executed or publicly disclosed between 2017 and 2023. This six-year window was selected to capture current market practice while providing sufficient historical depth to identify trends and variations.
B. Definitions Used
For purposes of this memorandum, the following terms are used consistently:
"License-Back" refers to a provision in a data-licensing agreement that grants the data licensor a license to use data, models, analytics, scores, or other outputs created by the licensee using the licensor's data as an input. License-back provisions vary widely in scope, duration, and conditions, and a central purpose of this survey is to catalog and analyze that variation.
"Derived Data" refers to data, models, scores, indices, analytics outputs, or enriched datasets created by the licensee using the licensor's raw data as an input. This definition is consistent with the definition of "Derived Data" as currently set forth in Section 1.14 of the draft MDLA under negotiation between Quorum and Meridian. The term encompasses a broad range of licensee-created outputs, from simple aggregations of raw data to complex multi-source blended models such as QuorumScore™.
"Revenue Share" refers to a percentage-based payment obligation tied to the licensee's commercialization of Derived Data to third parties. Revenue-share provisions, where present, typically require the licensee to pay the data licensor a specified percentage of net revenues (or, less commonly, gross revenues) generated from the sale, sublicensing, or other commercial distribution of Derived Data products.
These definitions are used throughout this memorandum unless otherwise indicated.
III. Overview of Meridian's Proposed License-Back and Revenue-Share Terms
A. The Three Components
Meridian's proposed license-back and revenue-share package, as set forth in Meridian's redline of Article VII of the draft MDLA (circulated by Graham Whitley of Kessler Pruitt & Ashe LLP on March 8, 2023), consists of three interlocking components:
1. Perpetual, Royalty-Free License-Back. Under proposed Section 7.2(a) of the MDLA, Meridian would receive a perpetual, irrevocable, royalty-free, non-exclusive, worldwide license to use, reproduce, distribute, display, perform, and create derivative works from all Derived Data for Meridian's own internal and commercial purposes. The license-back is not limited to any particular use case, product line, or business unit within Meridian. It expressly survives termination or expiration of the MDLA for any reason, meaning that Meridian would retain access to Quorum's Derived Data even after the parties' commercial relationship ends. The proposed provision contains no restrictions on Meridian's ability to sublicense the Derived Data to its own customers or partners, and no limitations on Meridian's use of the Derived Data in competition with Quorum.
2. Commercialization Gating with Sole Discretion. Under proposed Section 7.3(b), Quorum would be prohibited from selling, sublicensing, distributing, or otherwise making available any Derived Data to any third party unless and until the Derived Data has been "sufficiently transformed" such that "no Meridian Licensed Data can be reverse-engineered, reconstructed, or otherwise derived" from the output. The critical feature of this provision is that Meridian retains "sole and absolute discretion" to determine whether any particular output meets the "sufficient transformation" standard. The draft MDLA does not define "sufficient transformation" by reference to any objective technical criteria, does not provide for independent expert review, and does not include any appeal or dispute-resolution mechanism if Quorum disagrees with Meridian's determination. Quorum would be required to submit each proposed commercial output to Meridian for review at least sixty (60) days prior to any planned distribution, and Meridian would have no obligation to provide reasons for any determination that the output is insufficiently transformed.
3. 15% Revenue Share. Under proposed Section 7.4, if Quorum receives Meridian's written approval to commercialize Derived Data to third parties (i.e., if Meridian determines that the output is "sufficiently transformed"), Meridian would receive 15% of "Net Revenues" attributable to such Derived Data. "Net Revenues" is defined as gross revenues from the commercialization of approved Derived Data, less only direct third-party distribution costs (excluding Quorum's internal development costs, overhead, personnel costs, or marketing expenses). The revenue-share obligation would continue for the term of the MDLA and for a period of three (3) years following termination or expiration. The proposed provision does not include any cap on aggregate revenue-share payments, any sunset or step-down mechanism, or any methodology for attributing revenues to Meridian's data inputs versus Quorum's independent contributions in multi-source blended products.
B. Context — Meridian's Stated Rationale
Meridian's negotiating team has advanced several arguments in support of its license-back and revenue-share demands. As articulated by Arthur Pennington (General Counsel) and Keith Marlow (SVP, Data Licensing & Partnerships), and through Meridian's outside counsel at Kessler Pruitt & Ashe LLP (Graham Whitley), Meridian's position rests on the following premises:
intellectual Property Argument. Meridian contends that Derived Data constitutes a "derivative work" under copyright law (17 U.S.C. § 103) that remains subject to Meridian's underlying intellectual property rights in the raw licensed data. On this view, Quorum cannot create Derived Data without utilizing Meridian's copyrightable compilations and databases, and any output that incorporates or builds upon Meridian's data retains a nexus to Meridian's IP that justifies a license-back. We note that this legal characterization is contested — the copyright status of factual data compilations is limited under Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991), and the "derivative work" analysis for data-driven analytics outputs is unsettled — but a full analysis of the copyright issues is beyond the scope of this market-survey memorandum.
Data Moat and Valuation Narrative. Behind Meridian's formal IP argument lies a commercial and investor-relations motivation. Breckenridge Capital Partners Fund IV, L.P., which holds a 62.5% ownership interest in Meridian (with Caleb Rourke as the partner overseeing the Meridian investment), and Whitmore Zanthor & Co. (with Rachel Ng-Castillo as Managing Director advising Meridian on a potential 2025–2026 initial public offering) have expressed concern that granting broad derived-data rights to licensees like Quorum could undermine Meridian's "data moat" narrative. In the context of a pre-IPO company, the ability to demonstrate that downstream derivative uses of Meridian's data remain within Meridian's commercial orbit — through license-backs, revenue shares, and commercialization controls — supports a higher enterprise valuation by suggesting that Meridian's data generates value not only through direct licensing fees but also through participation in downstream analytics economics. This valuation narrative consideration, while commercially understandable from Meridian's perspective, does not reflect established market norms for data licensing and should not be allowed to distort the contractual terms of the MDLA.
Board Resolution. We note for completeness that Meridian's board of directors formally resolved, on August 22, 2023, against granting unrestricted derived-data commercialization rights in any data-licensing transaction. Although this resolution postdates both the period of our survey and the current stage of negotiations, the commercial dynamic that produced it was evident in earlier negotiations and informs Meridian's posture throughout. The resolution suggests that Meridian's position on license-back terms is unlikely to change absent significant counterpressure from Quorum.
IV. Market Survey Findings — License-Back Provisions
A. Prevalence of License-Back Provisions
Of the approximately thirty-five (35) enterprise data-licensing agreements surveyed, approximately twelve (12) — roughly 34% — contained any form of license-back provision relating to licensee-created derived data. The remaining approximately twenty-three (23) agreements (66%) contained no license-back whatsoever. In those agreements, the licensee retained sole and exclusive ownership of all derived outputs — including models, scores, indices, enriched datasets, and blended analytics products — subject only to standard restrictions on reverse engineering of the licensor's raw data, re-identification of de-identified data, and redistribution of the licensor's raw or unmodified data.
This finding is significant. It establishes that license-back provisions are a minority practice in enterprise data licensing, not a market standard. The majority of data licensors in the surveyed population treated their compensation as fully captured by the base license fee and did not seek additional rights in licensee-created derived outputs. In the majority-practice agreements, the parties typically addressed the derived-data issue through a combination of: (i) a clear ownership clause vesting all IP in derived outputs in the licensee; (ii) restrictions on redistribution of the licensor's raw data in unmodified or minimally modified form; (iii) prohibitions on reverse engineering of the licensor's data from derived outputs; and (iv) in some cases, non-compete or non-solicitation provisions restricting the licensee from using the licensor's data to build competing data products that would directly compete with the licensor's core data offerings. This majority-practice framework draws a clear distinction between the licensor's proprietary rights in its raw data (which are protected) and the licensee's proprietary rights in its derived outputs (which are respected), without requiring any license-back mechanism.
Conclusion: The existence of any license-back provision in the Quorum-Meridian MDLA would place the agreement in the minority of observed transactions. Meridian's proposal, which demands a broad and commercially unrestricted license-back, is further outside the mainstream.
B. Scope and Character of License-Back Provisions Where Present
Among the twelve (12) agreements that contained some form of license-back provision, the scope and character of the license-back varied substantially. We observed two broad categories:
1. Narrow, Internal-Use-Only License-Backs (8 of 12 Agreements)
The predominant form of license-back — found in eight (8) of the twelve (12) agreements containing any license-back — was a narrow, internal-use-only license limited in scope, purpose, and duration. These provisions shared the following common characteristics:
• Limited Subject Matter. The licensor received a non-exclusive, non-transferable license to use only aggregated or anonymized derived outputs, such as aggregate statistical models, benchmarking indices, or de-identified performance metrics. The license-back did not extend to the licensee's specific derived data products, proprietary models, algorithms, scoring methodologies, or enriched datasets. In several agreements, the license-back was expressly limited to "aggregate statistical outputs that do not identify or permit identification of any individual data subject or any specific output of [licensee's] proprietary models."
• Internal-Use-Only Restriction. The licensor was permitted to use the licensed-back outputs solely for its own internal purposes, typically limited to product improvement, research and development, benchmarking, and quality assurance. The licensor was expressly prohibited from using the licensed-back outputs for commercial purposes, including incorporating them into the licensor's own products or services offered to third parties, distributing them to the licensor's other customers, or using them for competitive purposes.
• Limited Duration. In most cases, the license-back was not perpetual. Six (6) of the eight (8) narrow license-backs were coterminous with the term of the underlying data license agreement, terminating upon expiration or termination of the main agreement. The remaining two (2) included defined survival periods of two to three years post-termination, after which the licensor's right to use the licensed-back outputs expired.
• No Sublicensing Right. None of the eight (8) narrow license-backs granted the licensor any right to sublicense the licensed-back outputs to third parties.
• No Revenue-Share Linkage. None of these eight agreements also included a revenue-share provision. The license-back was treated as a self-contained mechanism, and the licensor's compensation was limited to the base license fee plus the limited license-back right.
As a representative example, one credit-data licensing arrangement in the survey involved a major credit bureau licensing raw credit data to a financial services company that used the data to build proprietary risk-scoring models. The license-back provision in that agreement granted the credit bureau a non-exclusive, non-transferable, internal-use-only license limited to "anonymized aggregate model performance metrics" — i.e., high-level statistical information about how the licensee's models performed (e.g., overall accuracy rates, distribution of scores across population segments) — for the sole purpose of "improving and enhancing [licensor's] own scoring products and methodologies." The license-back did not extend to the licensee's proprietary models, algorithms, individual scores, or enriched datasets. It was coterminous with the license term and contained no sublicensing right.
2. Broader License-Backs (4 of 12 Agreements)
Four (4) of the twelve (12) agreements containing license-back provisions included broader license-back grants that permitted the licensor to use licensee-created derived outputs for commercial purposes (not limited to internal use). However, even in these agreements — which represent the outer boundary of observed market practice — the license-back was subject to significant limitations, conditions, and reciprocal compensation that distinguish them sharply from Meridian's proposal:
• Negotiated Carve-Outs. In all four agreements, the license-back was subject to specific, negotiated carve-outs that excluded the licensee's most commercially sensitive derived outputs from the scope of the license-back. Common carve-outs included: (i) the licensee's proprietary models, algorithms, and trade secrets; (ii) outputs that incorporated data from sources other than the licensor (i.e., multi-source blended outputs); and (iii) outputs that the licensee designated as "competitively sensitive" pursuant to agreed criteria. These carve-outs ensured that the licensor's license-back right did not extend to the licensee's core proprietary analytics products.
• Reciprocal Compensation. In three of the four agreements, the broader license-back was not royalty-free. Instead, it was accompanied by reciprocal compensation flowing from the licensor to the licensee, structured as fee reductions (Discounts on the base license fee), expanded data access (the licensee received access to additional data sets or higher-frequency data feeds at no additional cost), or co-branding credits (the licensor agreed to co-brand certain products utilizing the licensed-back data, with the licensee receiving attribution and co-marketing benefits). The fourth agreement provided that the licensor would pay a separately negotiated fee for any commercial use of the licensed-back outputs, determined on a case-by-case basis. In no case was the broader license-back granted on a purely royalty-free basis.
• Joint or Expert Determination. In all four agreements, the scope of the license-back — including what constituted "derived data" subject to the license-back and what fell within the carve-outs — was determined through a jointly agreed methodology or, in the event of disagreement, by an independent expert. No agreement granted the licensor sole discretion to determine the scope of the license-back or the eligibility of particular outputs.
• Duration Limitations. Two of the four broader license-backs were coterminous with the license term. The other two included defined survival periods (one year and three years, respectively) post-termination. None was perpetual with respect to specific models, scores, or enriched datasets.
Critical Finding: No surveyed agreement granted the licensor a perpetual, royalty-free, unrestricted commercial license to all licensee-created derived data without carve-outs, reciprocal compensation, or scope limitations determined by objective criteria or independent review. Meridian's proposed license-back is broader in scope, longer in duration, and freer of conditions than any license-back provision observed in the survey.
C. License-Back Duration
Duration is a critical dimension of license-back provisions, as it determines whether the licensor's rights in the licensee's derived outputs extend beyond the parties' commercial relationship. Among the twelve (12) agreements containing license-back provisions, the duration breakdown was as follows:
Perpetual license-back provisions were rare in the surveyed population. In the two agreements where a perpetual license-back element was present, the perpetual right was strictly limited to fully anonymized aggregate statistical data — e.g., aggregate benchmarking indices or high-level population-level statistical outputs — that could not be attributed to any individual data subject or used to reconstruct the licensee's proprietary models or specific derived products. These perpetual elements were understood by the parties as granting the licensor a right to retain high-level market intelligence derived from the licensing relationship, not as granting the licensor ongoing access to the licensee's proprietary analytics outputs.
Conclusion: Meridian's demand for a perpetual license-back covering all Derived Data — including proprietary models, scores, enriched datasets, and multi-source blended outputs like QuorumScore™ — is outside the observed range of market practice. Even the small number of agreements with any perpetual license-back element limited that element to anonymized aggregates, a far cry from Meridian's proposal.
D. Licensor Discretion Over Commercialization
Meridian's proposal to retain "sole and absolute discretion" to determine whether Quorum's derived outputs are "sufficiently transformed" for third-party commercialization is, in our assessment, the most anomalous element of the license-back package. This section reports our findings on commercialization-gating mechanisms in the surveyed agreements.
No surveyed agreement granted the data licensor sole discretion to determine whether licensee-created outputs were "sufficiently transformed" for third-party commercialization. Where transformation, de-identification, or independence-from-raw-data standards were included as a condition of the licensee's right to commercialize derived outputs — which occurred in approximately eight (8) of the thirty-five (35) agreements — the determination was made through one of the following mechanisms:
(i) Objective, measurable technical criteria specified in the agreement (5 agreements). These agreements defined "sufficient transformation" or its equivalent by reference to specific, quantifiable standards — for example, requiring that the derived output not permit reconstruction of any individual data point from the licensor's raw data with a specified confidence level, or requiring that the derived output incorporate data from at least a specified number of independent sources constituting at least a specified percentage of total data inputs by volume. These criteria were set forth in the agreement (often in a technical schedule or exhibit) and were testable by either party without requiring the licensor's subjective judgment.
(ii) A jointly appointed independent technical expert (2 agreements). These agreements provided that any dispute over whether a derived output met the applicable transformation standard would be resolved by an independent technical expert jointly selected by the parties (or, if the parties could not agree on an expert, appointed by a designated neutral institution). The expert's determination was binding on both parties. This mechanism ensured that neither party could unilaterally block the other's commercialization efforts.
(iii) Mutual agreement of the parties, with escalation to an independent expert (1 agreement). This agreement required the parties to attempt to reach mutual agreement on whether the derived output met the applicable standard, with escalation to an independent expert if mutual agreement could not be reached within a specified period (in this case, thirty (30) days).
In none of the surveyed agreements did the licensor possess a unilateral veto over the licensee's commercialization of derived outputs. The absence of sole-discretion gatekeeping in the surveyed population is not surprising. Industry commentators and practitioners have consistently identified sole-discretion standards as commercially impracticable and potentially abusive in the data-licensing context. Where the data licensor also operates in downstream analytics or scoring markets — as Meridian does — sole discretion to approve or reject the licensee's commercial outputs creates an inherent conflict of interest: the licensor has an incentive to withhold approval in order to prevent the licensee from competing in downstream markets, regardless of whether the output is genuinely "insufficiently transformed." This dynamic is sometimes referred to as "hold-up risk" and is widely recognized as a significant commercial concern in data-licensing negotiations.
The practical effect of Meridian's sole-discretion proposal would be to give Meridian an effective veto over Quorum's downstream commercial strategy. Because "sufficient transformation" is undefined in the MDLA and Meridian would have no obligation to provide reasons for a negative determination, Quorum would have no ability to predict whether any particular output would be approved, no ability to challenge a negative determination, and no recourse in the event that Meridian exercises its discretion in a manner designed to protect its own competitive position rather than to address legitimate concerns about the integrity of its raw data. This risk is particularly acute given the competitive overlap between Meridian and Quorum in downstream analytics markets.
Conclusion: Sole-discretion commercialization gatekeeping by the data licensor has no precedent in the surveyed agreements and is inconsistent with market practice. The observed market mechanisms for addressing transformation and de-identification standards uniformly rely on objective criteria, independent expert review, or mutual agreement — not unilateral licensor determination.
V. Market Survey Findings — Revenue-Sharing Arrangements
A. Prevalence
Revenue-sharing provisions tied to the licensee's commercialization of derived data were a minority feature in the surveyed agreements. Approximately six (6) of the thirty-five (35) agreements surveyed — roughly 17% — included any form of revenue-sharing mechanism. The remaining twenty-nine (29) agreements (83%) allocated all commercial value from derived data entirely to the licensee, with the data licensor's compensation limited to the base license fee (whether structured as a flat annual fee, a per-record fee, a tiered usage fee, or a combination thereof).
This finding reinforces the conclusion reached in Part IV.A above regarding the minority status of license-back provisions. The majority of data licensors in the surveyed population did not seek to participate in the economics of their licensees' derived-data commercialization beyond the base license fee. The base license fee was understood by the parties as full compensation for the licensor's contribution of raw data, and the licensee's investment in developing derived analytics products — including the cost of proprietary data inputs, model development, algorithm design, testing, validation, and commercialization — was treated as the licensee's investment yielding the licensee's returns.
B. Revenue-Share Rates Where Present
Among the six (6) agreements that included revenue-sharing provisions, the observed rates were as follows:
The median observed revenue-share rate was approximately 5–7% of net revenues. The range extended from 3% at the low end to 12% at the high end. The highest observed rate of 12% applied in a single agreement where the licensor's raw data constituted the predominant input to the derived product — specifically, where the licensor's data represented more than 80% of total data inputs by volume and the licensee's independent contribution was limited to aggregation, formatting, and basic statistical analysis without substantial proprietary modeling or blending with independent data sources. This fact pattern is materially different from the Quorum-Meridian arrangement, in which Quorum blends Meridian credit data with its own extensive proprietary first-party claims, policy, and telematics data and applies proprietary machine learning models and algorithms to produce QuorumScore™ outputs.
No surveyed agreement imposed a revenue-share rate of 15% or higher. Meridian's proposed 15% rate exceeds the top of the observed range by a margin of three percentage points and is approximately double the median rate. Even if one were to accept that some revenue share is appropriate in the Quorum-Meridian context (a proposition that we do not concede), the 15% rate demanded by Meridian is not supported by market practice.
C. Structure of Revenue-Share Provisions
Beyond the headline rate, the structural features of revenue-share provisions in the surveyed agreements are relevant to assessing Meridian's proposal. In all six (6) agreements containing revenue-share provisions, the revenue share was subject to conditions and limitations that are absent from Meridian's proposed terms:
• Caps. Five of the six agreements included caps on aggregate revenue-share payments, structured either as absolute dollar amounts (e.g., total revenue-share payments not to exceed a specified dollar amount per year or over the term of the agreement) or as a percentage of the total license fee (e.g., aggregate revenue-share payments not to exceed a specified percentage of the cumulative base license fee paid by the licensee). These caps ensured that the revenue-share obligation remained proportionate to the overall economics of the licensing relationship and did not grow to dominate the licensee's cost structure.
• Sunset Provisions. Four of the six agreements included sunset provisions under which the revenue-share rate declined over time or terminated altogether after a defined period, typically three to five years from the commencement of the licensee's commercialization of derived data. The rationale for sunset provisions is straightforward: over time, the licensee's independent investment in developing and enhancing the derived product increasingly dominates the licensor's initial data contribution, and the revenue share should diminish accordingly.
• Attribution Methodology. All six agreements included a clear, agreed-upon methodology for attributing revenues to derived data incorporating the licensor's inputs versus revenues attributable to the licensee's independent contributions. In multi-source blended products — such as a product that combines data from multiple licensors with the licensee's own proprietary data — the attribution methodology determined what portion of total revenues was subject to the revenue share. Common approaches included: (a) allocation based on relative data-input volumes; (b) allocation based on an agreed model-contribution analysis; and (c) allocation based on a jointly conducted attribution study using statistical techniques to isolate the marginal contribution of each data input.
• Good-Faith-Efforts Obligations. The revenue-share provisions were structured as obligations to pay a percentage of actual revenues, not as guaranteed minimum payments. The licensee was not obligated to achieve any particular level of commercialization revenues and was subject only to good-faith-efforts obligations to pursue commercialization opportunities.
Two additional structural features merit emphasis:
First, in all six (6) agreements with revenue-share provisions, the licensee retained sole ownership of the intellectual property in the derived data. The revenue share was characterized as additional consideration for the underlying data license — i.e., an incremental fee reflecting the licensor's contribution to the licensee's downstream commercial success — not as evidence of licensor co-ownership of, or co-authorship interest in, the derived outputs. This distinction is important because it confirms that a revenue share, even where present, does not imply or create any ownership interest by the licensor in the licensee's derived products.
Second, no surveyed agreement combined a perpetual royalty-free license-back with a revenue share. These two mechanisms serve overlapping commercial purposes: each is a means by which the licensor participates in the value created through licensee-derived outputs. The license-back gives the licensor direct access to use the derived outputs; the revenue share gives the licensor a financial stake in the licensee's commercialization of the derived outputs. Combining both mechanisms would effectively double-count the licensor's claimed interest, permitting the licensor to both use the derived outputs for its own commercial benefit (through the license-back) and extract a financial toll on the licensee's use of the same outputs (through the revenue share). This combination was not observed in any of the surveyed agreements, and its commercial logic is difficult to justify.
D. QuorumScore™ Implications
The implications of Meridian's revenue-share proposal for Quorum's planned QuorumScore™ B2B licensing business merit specific attention.
If Meridian's proposed 15% revenue-share rate were applied to Quorum's planned third-party licensing of QuorumScore™ outputs, the financial impact would be material. At the projected revenue level of $40–50 million in annual B2B licensing revenues by 2027, the revenue share would amount to approximately $6.0–7.5 million annually payable to Meridian. This sum is substantial both in absolute terms and relative to the overall economics of the Quorum-Meridian licensing relationship. For comparison, the base license fee under the proposed MDLA is $60 million over the five-year term ($10.8 million in Year 1 and $12.3 million in each of Years 2 through 5). A $6–7.5 million annual revenue-share payment would represent approximately 50–61% of the annual base license fee — effectively transforming Meridian from a data supplier into a de facto revenue partner in Quorum's downstream analytics business, notwithstanding that Quorum bears all of the development cost, commercialization risk, and competitive effort associated with building and marketing QuorumScore™.
Under the surveyed market practice, a blended model like QuorumScore™ — which combines Meridian credit data with Quorum's proprietary first-party claims, policy, and telematics data, and applies Quorum's proprietary machine learning models and algorithms — would typically be treated as predominantly a licensee creation. The data licensor's contribution of raw credit data would be addressed through the base license fee (here, $60 million over five years), and the licensee would retain all commercial rights in the blended output without further payment obligation. Even in the minority of agreements that included revenue-share provisions, carve-outs for the licensee's proprietary models and multi-source blended outputs were common. Several agreements expressly excluded from the revenue-share obligation any product in which the licensor's data constituted less than a specified percentage (ranging from 30% to 50%) of total data inputs, reflecting the view that where the licensee's independent contribution is substantial, the base license fee adequately compensates the data supplier.
The application of a 15% revenue share to QuorumScore™ — a product in which Quorum's independent first-party data and proprietary modeling constitute a significant (and arguably predominant) share of the total input value — is inconsistent with the structural logic observed in the surveyed agreements and would impose a disproportionate cost on Quorum relative to Meridian's actual data contribution.
VI. Market Survey Findings — Post-Termination Treatment of Derived Data
A. Observed Approaches
The treatment of licensee-created Derived Data upon termination or expiration of the data-licensing agreement is a critical commercial term and is currently a disputed point in the Quorum-Meridian MDLA negotiation. Our survey identified three approaches, with varying frequency:
Most Common Approach — Licensee Retains Derived Data (approximately 20 of 35 agreements). The most common approach, observed in approximately twenty (20) of the thirty-five (35) surveyed agreements, permitted the licensee to retain ownership of and continue to use Derived Data following termination or expiration of the agreement, subject to two conditions: (i) all raw licensor data must be returned to the licensor or destroyed within a specified period (commonly thirty (30) to ninety (90) days) following the effective date of termination, with the licensee providing a certified officer's certificate confirming return or destruction; and (ii) the Derived Data retained by the licensee must not contain or permit reconstruction of the licensor's raw data. This approach reflects the principle that the licensee's investment in creating Derived Data — including development costs, proprietary model design, and integration of independent data sources — generates an asset that belongs to the licensee and should not be forfeited upon termination of the data supply relationship. The licensor's interest in protecting its raw data is addressed through the return/destruction obligation, not through forfeiture of the licensee's derived outputs.
Less Common Approach — Continued Use with Wind-Down (approximately 10 of 35 agreements). In approximately ten (10) of the surveyed agreements, the licensee was permitted to retain Derived Data following termination, but subject to continued compliance with use restrictions and confidentiality obligations from the agreement, together with a defined wind-down period (typically six (6) to twelve (12) months) for transitioning away from commercial use of the Derived Data. During the wind-down period, the licensee could continue to use Derived Data for existing commercial commitments (e.g., fulfilling outstanding customer contracts) but could not enter into new commercial arrangements involving the Derived Data. At the end of the wind-down period, the licensee's right to use the Derived Data for commercial purposes terminated, although the licensee was typically permitted to retain the Derived Data for archival, regulatory compliance, and audit purposes.
Rare Approach — Destruction of Derived Data (approximately 5 of 35 agreements). In approximately five (5) of the surveyed agreements, the licensee was required to destroy Derived Data along with raw licensed data upon termination. This approach was observed primarily in highly regulated contexts — most notably, agreements governed by HIPAA requirements for protected health information and agreements involving classified government data — and was not observed in any credit-data or financial-data licensing arrangement comparable to the Quorum-Meridian transaction. In these agreements, the destruction requirement was driven by regulatory mandates rather than by the parties' commercial preferences.
Relevance to the Quorum-Meridian Negotiation. The MDLA currently treats the post-termination treatment of Derived Data as a disputed term. Meridian's position is that all Derived Data must be returned to Meridian or destroyed upon termination, treating Derived Data the same as raw licensed data for purposes of the post-termination return/destruction obligation. Quorum's position is that it should retain the right to continue using Derived Data post-termination, subject to return or destruction of all raw Meridian data within ninety (90) days. The market survey findings support Quorum's position: the majority approach (approximately 57% of surveyed agreements) permits the licensee to retain and continue using Derived Data post-termination, and the destruction-of-derived-data approach was limited to regulatory contexts not applicable here.
B. Interaction with License-Back
The interaction between license-back provisions and post-termination treatment of Derived Data raises a distinct concern specific to Meridian's proposal. In the small number of surveyed agreements that contained both a license-back provision and provisions addressing post-termination treatment of Derived Data, the license-back typically terminated upon expiration of the main agreement, consistent with the general pattern of coterminous license-back provisions described in Part IV.C above. The limited exception was for license-backs covering fully anonymized aggregate statistical data, which in two cases survived termination on a perpetual basis (as noted above).
Meridian's proposal creates a significant asymmetry. If the perpetual license-back demanded by Meridian is combined with Meridian's separate demand that Quorum destroy its Derived Data upon termination, the result would be a one-sided outcome in which:
(a) Meridian retains perpetual access to all Derived Data created by Quorum during the term of the MDLA, including QuorumScore™ outputs, for Meridian's own internal and commercial purposes; while
(b) Quorum loses access to its own derived creations — the products of Quorum's substantial investment in proprietary data, model development, and analytics capabilities — and is required to destroy them.
This outcome would effectively transfer the value of Quorum's derived-data portfolio to Meridian at no cost. It would mean that Meridian, following termination, could continue to use and commercialize outputs that Quorum developed using Quorum's own proprietary inputs and expertise, while Quorum itself would be prohibited from doing so. This asymmetry was not observed in any surveyed agreement and is, in our assessment, commercially unreasonable.
VII. Analytical Framework — Factors Driving License-Back Terms
The market survey findings reveal that the presence, scope, and terms of license-back provisions in enterprise data-licensing agreements are influenced by several identifiable factors. Understanding these factors provides context for assessing the reasonableness of Meridian's demands and for positioning Quorum's negotiating arguments.
A. Relative Contribution of Inputs
The most significant factor influencing the prevalence and scope of license-back provisions is the relative contribution of the licensor's data versus the licensee's independent inputs to the creation of derived outputs.
Agreements in which the licensor's data constituted the dominant input — by volume, by analytical significance, or both — to the derived outputs were more likely to include license-back provisions, and those provisions tended to be broader in scope. Conversely, where the licensee combined the licensor's data with substantial independent proprietary data, models, or algorithms, license-back provisions were less common and, where present, more narrowly scoped. This pattern reflects a commonsense intuition: the greater the licensee's independent contribution, the less justified is the licensor's claim to rights in the resulting output.
Application to Quorum-Meridian. QuorumScore™ is a multi-source blended model that combines Meridian credit data with Quorum's own proprietary first-party data sets — including historical claims data accumulated over more than two decades, policy-level underwriting data, and telematics data collected from Quorum's policyholder base — as well as Quorum's proprietary machine learning models and algorithms developed by Quorum's internal data science team. Quorum's independent contribution to QuorumScore™ is substantial by any measure. This fact pattern favors a narrower (or no) license-back under the analytical framework observed in the market, and it undercuts Meridian's assertion that its credit data contribution justifies a broad, unrestricted license-back on all Derived Data.
B. Competitive Dynamics
The competitive relationship between the licensor and the licensee is a second important factor. License-back provisions were more commonly observed in agreements between parties that operated in different market segments with limited competitive overlap. In those settings, granting the licensor a license-back to derived outputs posed less risk to the licensee's competitive position, because the licensor was unlikely to use the derived outputs to compete directly with the licensee's products or services.
By contrast, where the licensor competed with the licensee in downstream analytics or scoring markets, license-back provisions were rare and, where present, were subject to strict use restrictions and competitive-firewall obligations designed to prevent the licensor from using the licensee's derived outputs to gain a competitive advantage. Such restrictions included: prohibitions on the licensor's use of licensed-back data in any product or service that competed with the licensee's offerings; organizational firewalls within the licensor requiring that licensed-back data be accessible only to designated personnel who did not work on competing products; and audit rights allowing the licensee to verify compliance with competitive-use restrictions.
Application to Quorum-Meridian. Meridian operates in the credit-data analytics and scoring market, offering its own credit-scoring products and analytics services to financial institutions, insurers, and other enterprises. Quorum's planned B2B commercialization of QuorumScore™ and related analytics products will position Quorum as a competitor to Meridian in certain downstream analytics markets, particularly in the insurance and underwriting analytics space. This competitive overlap counsels strongly against granting Meridian a broad license-back, and it renders the sole-discretion commercialization-gating mechanism especially problematic. A provision that gives a competitor sole discretion to approve or reject Quorum's commercial outputs is not merely unusual — it is a structural competitive advantage that no rational licensee would grant absent extraordinary countervailing consideration.
C. IPO and Sponsor Dynamics
A third factor — less frequently discussed in industry commentary but observable in the surveyed data — relates to the ownership structure and capital-markets trajectory of the data licensor. In situations where a data licensor was backed by a private equity sponsor and preparing for an initial public offering, license-back demands were sometimes driven by the licensor's desire to demonstrate to prospective public investors that its data "moat" extends to downstream derivative uses. The ability to point to license-back provisions, revenue-share arrangements, and commercialization controls in the licensor's customer agreements supports a narrative that the licensor's data generates value not only through direct licensing fees but also through participation in the broader ecosystem of derived analytics. This narrative, in turn, supports a higher enterprise valuation and a more favorable IPO pricing.
While this motivation is commercially rational from the licensor's perspective — and we do not suggest that Meridian or its advisors are acting improperly in pursuing it — it is important to recognize that these demands are driven by the licensor's valuation narrative, not by established market norms for data-licensing terms. The cost of supporting the licensor's IPO story should not be borne by the licensee in the form of contractual terms that exceed market practice and impair the licensee's commercial position.
Application to Quorum-Meridian. This dynamic appears to be directly at play in the Quorum-Meridian negotiation. Breckenridge Capital Partners Fund IV, L.P. holds a 62.5% ownership interest in Meridian, and Caleb Rourke (the Breckenridge partner overseeing the Meridian investment) has been actively involved in the commercial terms of the MDLA negotiation. Whitmore Zanthor & Co., with Rachel Ng-Castillo as Managing Director, is advising Meridian on a potential 2025–2026 IPO. The connection between Meridian's aggressive license-back demands and the IPO valuation narrative is, in our assessment, not coincidental. Quorum should be prepared to identify this dynamic in negotiations — not as an accusation, but as a basis for distinguishing Meridian's demands from genuine market practice and for framing Quorum's counterproposal as the commercially normative position.
VIII. Comparative Analysis — Meridian's Proposal vs. Market Practice
A. Summary Comparison Table
The following table summarizes the comparison between Meridian's proposed license-back and revenue-share terms and the market practice observed in the survey:
B. Narrative Assessment
The comparative analysis confirms that Meridian's proposed license-back package is materially outside market norms, both in each individual element and in the aggregate combination of elements.
Individual Element Analysis. Each component of Meridian's proposal, considered in isolation, exceeds the range of market practice observed in the survey:
• The existence of any license-back places the agreement in the minority of observed transactions (34%).
• The scope of the license-back — perpetual, royalty-free, unrestricted as to use, and covering all Derived Data without carve-outs — exceeds even the broadest license-backs observed in the survey, all of which were subject to carve-outs, use restrictions, or reciprocal compensation.
• The duration — perpetual, surviving termination — exceeds all but two observed license-backs, both of which applied their perpetual element only to fully anonymized aggregate statistics.
• The sole-discretion commercialization-gating mechanism has no precedent in any surveyed agreement. Every observed transformation or de-identification standard was determined by objective criteria, independent expert review, or mutual agreement — never by the sole discretion of the data licensor.
• The 15% revenue-share rate exceeds the top of the observed range (12%) and is approximately double the median rate (5–7%). Even the 12% high-water mark applied in a fact pattern (licensor data constituting >80% of inputs) that is materially distinguishable from the Quorum-Meridian arrangement.
Combination Analysis. Beyond the individual elements, the combination of these provisions is itself unprecedented. No surveyed agreement combined a perpetual royalty-free license-back with a revenue share. These mechanisms serve overlapping purposes, and their combination effectively imposes a double cost on the licensee: the licensor obtains both direct access to the derived outputs (through the license-back) and a financial toll on the licensee's commercialization of the same outputs (through the revenue share). When the sole-discretion gatekeeping mechanism is added, the combination becomes triply burdensome: Meridian would receive free access to Quorum's derived outputs, a 15% share of Quorum's commercialization revenues, and an unreviewable veto over whether Quorum may commercialize at all.
Most Anomalous Element. The sole-discretion commercialization-gating mechanism is, in our assessment, the most commercially problematic element of Meridian's proposal. While a license-back can be negotiated to a narrow and tolerable scope, and a revenue-share rate can be negotiated downward, the sole-discretion gatekeeping mechanism poses a structural risk that cannot be mitigated through rate adjustments: it gives Meridian an effective veto over Quorum's downstream commercial strategy, exercisable without constraint, without transparency, and without recourse. This mechanism transforms the licensing relationship from a data-supply arrangement into a quasi-joint-venture in which one party holds unchecked authority over the other's commercial activities — a dynamic that is fundamentally inconsistent with the arm's-length data-licensing model and that has no support in observed market practice.
IX. Recommendations
Based on the market survey findings and comparative analysis set forth above, we recommend the following negotiating positions for Quorum, structured as a primary position and two fallback positions in order of preference.
A. Primary Position: No License-Back, No Revenue Share
Consistent with the majority market practice observed in our survey — in which 66% of agreements contain no license-back and 83% contain no revenue share — Quorum should continue to assert that it owns all intellectual property in Derived Data, with unrestricted rights to commercialize Derived Data to third parties, and that no license-back or revenue share to Meridian is required.
This position rests on the following grounds:
1. Adequate Compensation Through Base License Fee. Meridian is adequately compensated for its contribution of raw credit data through the base license fee of $60,000,000 over the five-year term of the MDLA ($10.8 million in Year 1 and $12.3 million in each of Years 2 through 5). The base license fee reflects the market value of Meridian's data and represents full consideration for all rights granted to Quorum under the MDLA, including the right to use Meridian's data as an input to the creation of Derived Data. The Pinnacle Ridge Advisory LLC pricing benchmark study (delivered February 14, 2023, prepared by Franklin Osei) contains comparative data on license economics that supports this position.
2. Quorum's Substantial Independent Contribution. Quorum contributes substantial independent proprietary data (claims, policy, and telematics data accumulated over more than two decades), proprietary machine learning models and algorithms developed by Quorum's internal data science team, and significant development resources to the creation of Derived Data products, including QuorumScore™. This independent contribution is the primary source of value in Quorum's derived analytics products, and Quorum should own and control that value without encumbrance.
3. Market-Practice Support. The primary position is the majority-practice position in the surveyed market, adopted by approximately two-thirds of comparable agreements. It is the presumptive norm in enterprise data licensing, not an aggressive or outlier position.
4. Protection of QuorumScore™. The primary position protects Quorum's ability to develop, commercialize, and license QuorumScore™ outputs without restriction, preserving the projected $40–50 million B2B revenue line and the value of Quorum's pending trademark application (USPTO Serial No. 97/482,116).
B. Fallback Position: Narrow, Time-Limited License-Back Without Revenue Share
If a license-back is commercially necessary to close the MDLA transaction — and we recognize that Meridian has indicated through counsel that some form of license-back may be a condition to closing — Quorum should negotiate for a license-back provision that falls within the range of observed market practice, specifically:
(a) Scope — In

MEMORANDUM

Record

Parties

Dates

Amounts