# EU portfolio data referral eligibility | SourceX

Canonical: https://refer.sourcex.si/guides/eu-portfolio-data-referral-eligibility
Last updated: 2026-10-20

Before referring an EU portfolio company, an introducer should ask five screening questions: how many full-time staff the specific entity had at peak, how long its records go back, which systems hold them, who owns them, and who can authorize a licensing discussion. Those questions decide eligibility. Questions about GDPR lawful basis, purpose compatibility, client contracts and AI Act transparency belong to a separate rights review that SourceX runs with the company and its counsel after the company applies.

## What should an introducer ask before referring an EU portfolio company?

Ask questions you can answer from board packs, management conversations and the deal file, and stop there. The goal is to decide whether one named entity is worth an introduction, not whether its records can lawfully be licensed.

The most useful habit is to keep two lists apart. Mixing them is how portfolio introductions stall: a deal partner hears "GDPR" and either drops a good candidate or starts collecting documents nobody should be collecting.

### Gate one: eligibility screening (the introducer's job)

| Question | Why it matters | Where the answer usually is |
|---|---|---|
| Which legal entity holds the records? | Eligibility and ownership attach to the entity, not the fund or group | Structure chart in the deal file |
| Full-time employees at peak, contractors excluded? | The program needs 50+ | Board pack or HR summary |
| How many years of operating records, in which systems? | Documented operating history | Management conversation |
| Where is the entity, and for non-EU entities, is at least 90% of the data in English? | Geography rule below | Management conversation |
| Who can authorize a licensing discussion? | An authorized contact must apply | CEO, or an officer the board designates |

### Gate two: jurisdiction-specific rights review (not the introducer's job)

| Question | Who answers it | Why the introducer should not |
|---|---|---|
| Is there a lawful basis, and is licensing compatible with the purpose the data was collected for? | Company counsel and SourceX's review | Requires seeing the records and their collection context |
| Do client or supplier contracts restrict reuse? | Company counsel | Requires reading contracts the introducer should not hold |
| Which personal data must be removed or de-identified? | SourceX data preparation, with the company | Requires access to content |
| What will downstream AI developers need to disclose? | SourceX and the buyer | Depends on the buyer and the dataset |

The European Commission summarizes the GDPR's seven principles, including lawfulness, fairness and transparency, purpose limitation and data minimization, and the EDPB adds that controllers must be able to demonstrate compliance with them. That is precisely why gate two needs the controller (the portfolio company) at the table with the records in front of it. On the buyer side, the Commission's guidance says providers of general-purpose AI models must publish a summary of their training content using a mandatory template, which is another reason licensed data needs a clean rights record. Background is in the guides on [compatible further processing](/guides/gdpr-further-processing-compatible-purpose-ai-training), [the EU AI Act training-data summary](/guides/eu-ai-act-training-data-summary-template) and [excluding EU personal data from a dataset](/guides/exclude-eu-personal-data-from-dataset).

## Does "English records" decide eligibility for an EU company?

No. The published rule reads: 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 90% of the data is in English.

So the 90% English test applies to the second group, not to an EU entity. An EU company whose records are largely English is still a straightforward conversation, and an EU company whose records are mostly in German or French is not excluded by the geography rule; whether those records fit a particular dataset is a question for SourceX's review. Where the language test matters in a European portfolio is the sister company across the Channel: a UK entity needs at least 90% of its data in English.

## An illustrative portfolio screen

Illustrative. A fictional fund with four European holdings. No real companies, and no figure here is a forecast.

| Entity | Full-time at peak | Location | Records and span | Authority | Gate one result |
|---|---|---|---|---|---|
| Packaging manufacturer | 310 | Germany | Quality and maintenance logs, 11 years | CEO | Introduce |
| Software support provider | 120 | Ireland | Resolved tickets with outcomes, 8 years | CEO, with board approval | Introduce |
| Field services firm | 95 | UK | Job sheets and inspection reports, 9 years | Managing director | Confirm the 90% English share first |
| Regional holding company | 30 | Netherlands | Group finance only | Fund partner | Do not introduce: under 50 |

Notice what the table does not contain: no GDPR conclusion, no contract analysis, no sample. Two entities move forward, one needs a single factual answer, and one stops. Everything else waits for gate two.

## What should the fund partner avoid doing?

- Do not request exports, data rooms or "a quick sample" from portfolio management. Partners never handle raw client records.
- Do not apply on the company's behalf. The portfolio company decides, and an authorized person applies through your referral link.
- Do not promise an outcome. Until SourceX reviews the entity, nobody knows SourceX's current data needs for its records, the rights review result, the company's compensation or any reward.
- Do not use a valuation conversation as leverage. If a holding is under pressure, the guide on [what to do after a portfolio company markdown](/guides/portfolio-company-markdown-what-to-do) treats that decision separately.

For the CEO conversation itself, use the [questions to ask a portfolio company CEO about AI](/resources/questions-to-ask-portfolio-company-ceo-about-ai). For screening a U.S. portfolio, the guide for [private equity teams](/guides/how-private-equity-teams-can-assess-portfolio-company-data-opportunities) covers the same ground with U.S. examples.

## How are company compensation and the introducer's reward kept apart?

They are separate payments with separate bases. The portfolio company is paid under its own agreement with SourceX. Rewards are a share of the compensation SourceX actually pays the company you referred (the Actual Supplier Payout), up to $100,000 per referred company, subject to the program terms. SourceX pays rewards separately; they never reduce the company's payout. Credit goes to the first valid introduction within the 90-day window, recorded when the company submits a verified application, and one partner is credited per company. A fund that introduces three unrelated portfolio companies makes three separate introductions, each with its own per-company cap; a parent company and its subsidiaries, by contrast, count as one company with one cap across the group. The [rewards page](/rewards) explains how statuses move.

## Next step

> **Refer a company. Earn up to $100,000.** Default reward: 25% of Actual Supplier Payout, capped at $100,000 cumulatively per referred company. Eligibility and payment conditions apply, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.

When an entity clears gate one and its CEO agrees, [register as a partner](/register), confirm the details against [who qualifies](/who-qualifies), and keep the [program terms](/terms) on hand; gate two starts once the company applies.

## Sources

- https://refer.sourcex.si/terms
- https://commission.europa.eu/law/law-topic/data-protection/data-protection-explained_en
- https://www.edpb.europa.eu/topics/key-gdpr-concepts/basic-principles_en
- https://digital-strategy.ec.europa.eu/en/faqs/template-general-purpose-ai-model-providers-summarise-their-training-content