# Ownership blockers in referrals: a study design | SourceX

Canonical: https://refer.sourcex.si/guides/how-often-do-ownership-questions-stop-an-introduction
Last updated: 2026-11-04

Nobody can say yet. SourceX has published no figure for how often doubts about who owns a company's records, or about who controls the company, end an introduction, and this page reports none. It sets out how a platform research team would measure it: an eleven-code taxonomy in two families, a rule that a question counts as stopping an introduction only when a dated record says so, and a stop rate computed over every introduction in a cohort, with other endings treated as competing outcomes rather than dropped.

The figure matters because the word "ownership" hides two different problems. One can end an opportunity for good. The other usually changes whom to ask, and when.

## Whose records, and who decides: two questions under one word

SourceX's screening page puts the distinction in two sentences: "Ownership changes who says yes and where a check usually breaks. It never changes the checks themselves."

- **Record questions** ask whose records these are and whether the company may license them. The screening page counts a company as holding the rights when it owns the material, or can license it without breaking a promise of confidentiality to a client, an employee privacy commitment or a contract with someone else.
- **Control questions** ask who decides for the company today. Under the program the CEO authorizes a licensing discussion. A sale, a group structure, a formal insolvency process or deadlocked owners can change which CEO that is, or whether the CEO alone is enough.

A study that pools the two will report one rate that describes neither.

## The ownership blocker taxonomy

| Code | Family | The question as someone states it | Usual disposition | Who can answer it |
|---|---|---|---|---|
| R1 Client-owned | Records | The records are kept for clients and belong to them | Often stops: client material is usually not a fit | The company and its counsel |
| R2 Contract limit | Records | A customer, vendor or partner contract may restrict reuse | Resolved or stopped in review | The company's counsel and SourceX's rights review |
| R3 Held elsewhere | Records | The records sit in a system the company does not control | Often stops | The company |
| R4 Made outside | Records | Contractors or agencies produced the material and the company is unsure of its rights | Resolved or stopped in review | The company's counsel |
| R5 Personal data | Records | The records hold employee or other personal information | Resolved through scope, or stopped | The company's counsel and SourceX's review |
| R6 Licensed before | Records | The records were already licensed for AI training | Stops | The company |
| C1 Change of control | Control | A sale is signed or has closed | Pauses: the target's CEO decides before closing, the acquirer's CEO after | Whichever CEO is in charge at the time |
| C2 Group entity | Control | The records sit in a parent, subsidiary or sister company | Changes whom to ask | The CEO who decides for that entity; for a platform with add-ons, the platform CEO |
| C3 Formal process | Control | A receiver, trustee or assignee controls the assets | Changes whom to ask; the CEO alone may not be enough | Whoever controls the estate |
| C4 Deadlocked owners | Control | Two equal owners mean the CEO alone may not be enough | Pauses or stops | The owners and the CEO |
| U Unclassified | Either | An ownership doubt that fits no code | Recorded as stated | Nobody yet |

**Where each code comes from (methods note, 2026-10-28).** R1, R2, R3, R5 and C1 to C4 follow SourceX's screening page. R6 follows what a supplier confirms under SourceX's published process: the data "hasn't been licensed for AI training before". R4 follows the screening page's emphasis on material a company created itself.

**C2 and credit.** A parent and its subsidiaries form one company for the reward cap. Whether earlier contact between SourceX and a parent, or a different partner's introduction of a sister company, changes who is credited is something SourceX settles case by case in duplicate review. The study records the decision when the partner is told it, and never predicts one.

**Not ownership.** A failed size or history fact, a language mismatch, a program with no current need, a CEO who declines for other reasons, and silence. These endings stay in the cohort as competing endings.

## Raised, resolved and stopped are separate counts

Every ownership question gets four fields: its code, the stage at which it surfaced, who raised it and how it ended.

| Field | Values |
|---|---|
| Stage | S1 between the partner's first mention to the CEO and a submitted introduction; S2 between submission and an application; S3 during SourceX's review |
| Raised by | The partner, the company (CEO or its counsel), or SourceX |
| Disposition | Resolved and continued; stopped the introduction; still open at the read date; unknown |
| Record | The dated message or note the code rests on |

**No record, no stop.** A question counts as stopping an introduction only when a dated record names it as the reason: the CEO's own message, a statement from the company, or a SourceX stage note shared with the partner. A company that stops without saying why is coded "stopped, reason not recorded" and stays in the count under that heading.

**Several questions, one introduction.** An introduction can carry an R2 and a C1 at once. Each question is coded. The introduction is counted once in the overall ownership stop rate and once under each code the stopping record names.

## Measures, with other endings treated as competing

- **Raise rate.** Introductions with at least one ownership question, divided by every introduction in the cohort.
- **Ownership stop rate by day 180.** The cumulative incidence of a recorded ownership stop, with the other endings (a failed fit fact, no current program need, a CEO declining for another reason, 60 days of silence) treated as competing events. Companies still active at day 180 are counted as open.
- **Resolution share.** Questions resolved, divided by questions raised, for each code.
- **Where questions surface.** The split of questions by stage and by who raised them.

The competing-events rule is the one statistical choice that changes the headline. A tutorial in the American Heart Association's journal Circulation advises that "when estimating the crude incidence of outcomes, analysts should use the cumulative incidence function, rather than the complement of the Kaplan-Meier survival function." In plain terms: a company that stopped for some other reason can no longer be stopped by ownership, and a method that pretends it still could will overstate the rate.

## The coding protocol

- **Cohort.** Every company whose CEO hears about SourceX from a participating partner during two quarters, entered on the date of that first dated mention and followed to day 180. Here an introduction starts at that mention, submitted or not, so questions that end things before submission are counted too.
- **Sources.** The partner's dated log of what it was told; statements the company chose to make, with its consent; SourceX stage records shared with the partner. Nobody in the study reads a contract, a cap table, a record or a sample.
- **Blind coding.** Two coders assign codes from the record of each raise without seeing how the introduction ended. A third person, without seeing the codes, assigns dispositions from the final records. Disagreements are discussed and logged, and agreement for each code is published before any rate.
- **Pilot.** Before enrollment, code 20 past introductions with the partners' consent. If unclassified questions exceed one in ten, revise the codes and date the revision.
- **Consent and withdrawal.** Partners agree in writing. A company may ask to be removed; it is counted as withdrawn, never deleted from the total.

## Four introductions coded (Illustrative)

Illustrative. Generic companies and invented events; no outcome is implied.

| Introduction | Question and code | Raised by, at stage | Disposition | Counts toward |
|---|---|---|---|---|
| A 260-person facilities group | R2: a national customer's contract may limit reuse of service tickets | The company's counsel, S2 | Resolved: the company scoped its application to internal records | Raise rate and resolution share |
| A 140-person payroll bureau | R1: the payroll records belong to employer clients | The CEO, S1 | Stopped, recorded in the CEO's reply | Raise rate and ownership stop rate |
| A 90-person dental lab group | C1: a sale is signed, closing in two months | The CEO, S1 | Open at day 180: the CEO asked to leave it to the buyer's CEO after closing | Raise rate, open |
| A 180-person regional distributor | None | Not applicable | The CEO declined for a busy season | Denominator only, as a competing ending |

The distributor matters as much as the bureau. Leave it out and the bureau's stop looks like one in three; keep it in and the ownership stop is one ending among four.

## The reporting table, blank until day 180

| Figure | Numerator | Denominator | Status |
|---|---|---|---|
| Raise rate | Introductions with any ownership question | Every introduction in the cohort | Not yet published |
| Ownership stop rate by day 180 | Introductions stopped for a recorded ownership reason | Every introduction, other endings competing | Not yet published |
| Resolution share | Questions resolved, by code | Questions raised with that code | Not yet published |
| Stage and source | Questions by stage and by who raised them | All questions raised | Not yet published |
| Unrecorded stops | Stops with no recorded reason | All stops | Not yet published |

If unrecorded stops exceed a third of all stops, no ownership stop rate is published, only counts. Cells under five companies are folded into "other", and the analysis plan is dated before the first company enters.

## Results that would reshape the taxonomy

- **If R1 dominates at S1**, the problem is mostly settled in the first conversation, and a better question before the ask matters more than anything later.
- **If control codes mostly pause rather than stop**, calling them blockers overstates the problem, and the two families should be reported as separate studies.
- **If coders cannot tell R2 from R4**, the codes merge, and the published taxonomy says so.
- **If unrecorded stops are common**, the study is measuring record-keeping rather than ownership.

## Where an ownership study can mislead

- **Silent pre-screens.** A partner who never mentions SourceX because the records look like a client's leaves no trace; the rate describes asks that were made.
- **Statements, not documents.** Codes rest on what people said. Rights are settled in SourceX's review and by the company's counsel, which may read a contract differently.
- **Polite reasons.** "A contract issue" can be a courteous way to decline. The record is coded as given.
- **Moving criteria.** SourceX applies each program's own criteria, so the same records can meet a different answer later.

## Using the taxonomy before any rate exists

The codes work today as a vocabulary for one conversation. Before raising SourceX with a CEO, a partner can check the published fit facts: at least 50 full-time employees at the company's peak, contractors left out; years of records the company made itself; the right to license them; and a CEO who wants to talk. 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. Whether customer contracts allow reuse, and whether a program needs the records now, stay open until SourceX reviews them.

Questions stay at the level of a sentence. The [five-question ownership intake](/guides/how-to-ask-about-ownership-without-creating-a-diligence-project) shows where to stop, and the [system ownership questions](/resources/company-system-ownership-checklist) are for the company's own team once it is talking to SourceX. Partners sign the Channel Partner Agreement before submitting any introduction and never handle records. For the wider set of reasons introductions slow down, see the [stall study design](/guides/the-most-common-reasons-company-introductions-stall); for a score that counts rights questions alongside other friction, see the [friction index](/guides/the-data-introduction-friction-index-what-it-should-measure). Other designs are in the [guides hub](/guides).

## 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.

When a company's records are its own and the CEO in charge today wants the conversation, [register as a partner](/register), recheck it against [who qualifies](/who-qualifies) and read the [program terms](/terms) on duplicates before you submit.

Coding why introductions stop is network research, and deciding whether to invest in SourceX is a different question; anyone weighing it can request investor materials from the [SourceX investor overview](https://sourcex.si/investors).

This article is for information only and is not an offer to sell or a solicitation of an offer to buy securities. Any offering by SourceX is made only to verified accredited investors through its offering documents.

## Sources

- https://refer.sourcex.si/terms
- https://refer.sourcex.si/who-qualifies
- https://sourcex.si/how-it-works
- https://www.ahajournals.org/doi/10.1161/circulationaha.115.017719
- https://sourcex.si/investors
