# Enterprise data partnership models compared | SourceX

Canonical: https://refer.sourcex.si/guides/enterprise-data-partnership-models-compared
Last updated: 2026-11-04

Three structures sit behind the word "partnership". In a direct license, one counterparty contracts with the company, reviews rights, prepares and delivers the data and pays the company. In a sourcing program, the company connects tools or supplies records to a platform that runs its own pipeline and sets compensation by volume or value. In a marketplace, the company lists datasets and is paid per license as buyers come. Ask who signs, who prices and who the buyer is.

## How do direct company licensing, data sourcing and marketplace partnerships differ?

They differ in where three decisions sit: who the company contracts with, who sets the price, and how many buyers see the data. Everything an introducer tells an owner about effort, timing and control follows from those three answers.

The labels overlap in marketing copy, so read the mechanics rather than the name. A program called a "partnership" may be a single licensing agreement, a continuing data feed, or a listing on a catalog. The table below reads each program's own public pages, checked on 2026-10-18, and leaves a field as "not stated" where the page does not answer it.

## The model map: four programs, read from their own pages

None of the companies below is a SourceX buyer, client or partner. They are compared only on what their public pages say.

| Field | SourceX | Mercor (data) | micro1 (data partnerships) | Defined.ai (partnership programs) |
|---|---|---|---|---|
| Model | Direct license | Sourcing program | Sourcing program with acquisition budget | Marketplace |
| Company's counterparty | SourceX | Mercor | micro1 | Defined.ai, then a Data License Agreement |
| What the company supplies | Approved business records selected in review | Data from connected tools | Operational data such as SOPs and knowledge bases | Packaged text, audio, image, video or multimodal datasets |
| How company pay is set | Offer after review, paid by SourceX | Data volume, tools connected and depth of data | Characteristics and value of the data | Dataset type, usage and customer needs, approved by the partner |
| Who reaches buyers | SourceX, for the selected dataset | Mercor | micro1, which says it builds RL environments from de-identified data | Marketplace demand from many buyers |
| Company size signal | 50+ full-time employees at peak | Not stated | 30+ employees preferred | Not stated |
| Published company-referral reward | the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; the USD 50,000 floor applies once per company | Not published | $50K for enterprise data partnerships up to a deal-size threshold, then 7.5% of the deal value, with no published cap (checked 2026-10-26) | Not stated |

Three things stand out. First, only the marketplace model leaves the company facing many buyers: Defined.ai's supplier terms grant it a non-exclusive license to show samples to customers, and the commercial terms are agreed dataset by dataset. Second, the sourcing programs price by the data's characteristics or the tools connected rather than by a listing price. Third, micro1 has announced a commitment to spend $1 billion over 12 months on enterprise data acquisition; that is a statement of intent, not evidence that the money is spent or that a given company will be paid.

For the referral side of two of these programs in more depth, see [Mercor vs SourceX](/guides/mercor-vs-sourcex-data-referral) and [SourceX vs micro1](/compare/sourcex-vs-micro1-company-referral). For a structured way to score any program, use the [referral program scorecard](/guides/ai-data-referral-program-comparison-scorecard).

## What does each model ask the company to do?

The effort differs more than the headline money does. Illustrative: a 260-person insurance claims administrator with nine years of adjuster notes and claim workflows.

| Step | Direct license | Sourcing program | Marketplace |
|---|---|---|---|
| First conversation | Metadata only: what records, how many years, which system, who owns them | Usually an application and tool connection discussion | An application and dataset description |
| Rights work | Counterparty runs a rights review before an offer | Contractual protections described by the platform | Supplier provides evidence of lawful collection and completes vetting |
| Data preparation | Counterparty prepares and de-identifies the selected set | Platform pipeline processes connected data | Supplier packages the dataset to the catalog's standard |
| Pricing moment | One offer for a defined dataset | Ongoing, based on volume or value | Per license, as buyers appear |
| Who the company talks to | One counterparty throughout | One platform | The marketplace, under a separate data license |

The column that fits depends on the company. A claims administrator that has never packaged data will usually want one counterparty to do the preparation. A company that already sells datasets may prefer a catalog. The [marketplace vs managed licensing comparison](/compare/data-marketplace-vs-managed-data-licensing) and [selling directly vs through a marketplace](/compare/sell-data-directly-to-ai-companies-vs-marketplace) go further on that choice.

## Company compensation and referral rewards are separate numbers

Keep two figures apart in every conversation. The company's compensation is whatever its own agreement says. The introducer's reward is set by the referral program and, at SourceX, is calculated on what SourceX actually pays the company, funded separately, and never deducted from the company's payout.

Comparing programs on referral reward alone misses that link. A reward calculated on what the company is actually paid moves with the company's outcome; a fixed bonus, a revenue threshold or an unstated amount does not tell you how the two relate. Illustrative: if SourceX paid a company $250,000 for a completed handoff, the SourceX reward on that first qualifying transaction would be $62,500; at $800,000 it would be $120,000, because there is no cap and 15% of the payout is the larger figure at that size. micro1's company-referral page, checked 2026-10-26, pays $50K for enterprise data partnerships up to $700,000 and 7.5% of the deal value above that, with no published cap; Mercor's data page describes a bonus without a figure. Neither is a reason to steer an owner toward one model: the owner's choice should follow the effort and control table above.

## Which questions qualify the company for any of these models?

Whatever the model, you can screen at metadata level without touching records:

- Which systems hold the records, and for how many years?
- Did the company create them, and do customer contracts allow reuse?
- Is the material mostly business records rather than consumer personal data?
- Which executive can authorize a licensing discussion?
- Would the company rather have one counterparty prepare the data, or package it itself?

For a SourceX introduction specifically, the company needs 50+ full-time employees at peak (contractors excluded), a documented operating history, rights to license and an authorized contact. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 70% of the data is in English. [How to introduce a company for an AI data partnership](/guides/how-to-introduce-a-company-for-an-ai-data-partnership) covers the introduction itself.

## What stays unknown from public pages?

Public pages answer structure, not outcomes. Not published by any of the four: acceptance rates, typical company compensation for a given record type, time from application to payment, and how often datasets are re-licensed. If an owner needs those answers, the honest route is to ask each program in writing and compare the replies. Also unknown until review at SourceX: SourceX's current data needs for the specific records, the rights review result and the offer.

## Next step

> **Refer a company. No cap on rewards.** Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.

If the owner wants one counterparty to run rights review, preparation and delivery, [register as a partner](/register), check the company against [who qualifies](/who-qualifies), and read the [program terms](/terms); the [rewards page](/rewards) explains how payment is sequenced.

## Sources

- https://refer.sourcex.si/terms
- https://www.mercor.com/data/
- https://www.micro1.ai/data-partnerships
- https://www.micro1.ai/blog/micro1-commits-1b-to-enterprise-data-acquisition
- https://www.micro1.ai/company-referral
- https://www.micro1.ai/forum/company-data-partnerships-referral-program
- https://defined.ai/partnership-programs
- https://defined.ai/supplier-program