The study is the platform team's own. SourceX does not run it, and it needs nothing from SourceX beyond the stage your dashboard already shows for each referral.
Where this protocol sits beside the friction map
The friction map follows companies forward from a CEO's yes, with a journey log the company's contact keeps for 90 days. This protocol looks back. It interviews people after each journey has ended, covers asks that never reached a yes, and is built around one constraint the map treats as a column: the study holds nothing about any company's records. It reuses the map's twelve friction codes rather than inventing new ones. For how often a change of owner loses an introduction, see the handoff attrition study; for which measures an index should track, the friction index; and for a forward-looking stage clock with a blind reason codebook, where and why introductions stall.
Why friction studies drift toward company records
Ask why an introduction stalled and people explain with the company's facts: "their counsel found most of eleven years of tickets name customers," or "they only had 40 on payroll after the layoffs." The explanation is natural and the facts are not yours to keep. A screenshot of an application form, a forwarded email from the CEO or a pasted line from an inventory arrives the same way, offered to help.
SourceX's own pages set the baseline the study must stay under. Its process page tells partners: "You do not need to export, upload, or describe confidential records." The unsigned sample of the Channel Partner Agreement goes further: section 4.2 leaves the partner without authority to pass on a supplier's records or the personal information in them, and section 4.4 bars a partner from collecting or keeping copies of supplier datasets through the program unless separately authorized in writing. A study that quietly filled up with record detail would hold more than the referral itself ever needed.
The capture boundary
One page, agreed before the first interview, read aloud at the start of each one.
| Topic | Write down | Never write down | If volunteered, write instead |
|---|---|---|---|
| Timing | Weeks from first mention to the ask, the answer, any application and the last contact | Dates about the company's systems, such as go-live or retention years | "Speaker gave a system date; struck" |
| People | Roles: the CEO, the relationship owner, "a second senior person at the company" | Names of anyone at the company, the CEO included, and personal details about anyone | "Another company role joined in week 4" |
| Obstacles | The obstacle in the speaker's words, with company detail removed | The record facts behind the obstacle | "A question about whose records they were, raised by the company's counsel" |
| The records | Nothing | Record types, volumes, years, systems, samples, exports | "Speaker described the records; struck" |
| Fit facts | Only that a fit question arose, and which kind: size, history, rights or place | Headcount, founding year, revenue, customer names | "A size question arose in week 3" |
| SourceX | The stage your dashboard showed, and its date | The content of any confidential message, anything about Acquisition Partners | "Stage changed on the dashboard in week 7" |
| Documents | That one was offered and declined, or arrived and was deleted, with the date | Screenshots, forwards, attachments, form contents | "Document arrived unasked; deleted the same day" |
Peak headcount is the hard case. A partner may properly discuss it with a CEO when screening, but a friction study does not need it to explain a delay, so it stays out of the file.
The journey interview protocol
Sampling frame. Every company the team named for an ask during two fixed quarters, whatever happened next: asks never made, asks declined, yeses that went quiet, applications that went through. A journey enters the frame when it closes: a verified application, an explicit stop, or 120 days after the ask with no news.
Who is interviewed. Up to three seats per journey, each separately: the relationship owner, the platform lead, and the company's CEO or a person the CEO names, only if the CEO agrees. Interviews happen within three weeks of a journey closing, while memory is fresh.
Consent. The Belmont Report treats informed consent as three elements, information, comprehension and voluntariness (our paraphrase). Each one has a concrete step here:
- Information: the interviewee hears the study's purpose, the capture boundary, who will see results (aggregates only) and when notes are deleted.
- Comprehension: the interviewer asks them to say back, in their own words, what the study never writes down.
- Voluntariness: they may decline, skip a question or stop. A company that declines loses nothing, and the fund tells SourceX nothing about who took part.
Six prompts, in timeline order.
- "When did this company first come up, and who raised it?"
- "Who asked the CEO, how, and on roughly what date?"
- "What happened between the ask and the answer? Was there any wait longer than two weeks?"
- "Where did it slow down or stop, in your words? I will note the kind of question, not the company's details."
- "Who held the next step at each point?"
- "If our side did this again, what would you change?"
Company-side interviews add two: "What did you need that you did not get?" and "Was anything unclear about who decides, who pays whom, or what you were asked to share?"
The strike rule. When a speaker volunteers record detail, the interviewer says "I'll leave that detail out," writes the substitute line from the boundary table and moves on. A document offered is declined; one that arrives anyway is deleted the same day and logged as a strike.
Notes, not audio. The interviewer types notes live, the interviewee reviews them within five working days and may correct or strike anything, and only then are they coded. Raw notes are deleted once coding is checked, on a date stated in the consent.
Variables and the proposed analysis
| Variable | Values |
|---|---|
| Journey code | A number; the key to company names is held by one person and destroyed when the report is published |
| Seats interviewed | Owner, platform lead, company side; each yes, declined or not asked |
| Furthest stage | Named for an ask; asked; CEO answered; applied; application verified |
| Weeks between stages | From the interviews, checked against the team's own dated records |
| Friction codes | The friction map's twelve codes, applied by two coders working apart |
| Next-step holder | Who held it at each stage, by role |
| Strikes | Count of struck details and declined or deleted documents |
Analysis. Counts by furthest stage and by code, with every journey in the frame in every denominator, interviewed or not. Where two seats disagree on a date by more than a week, or on the reason a journey stopped, the disagreement is reported in its own table rather than settled. Every finding is descriptive: a code that shows up more often in stalled journeys is associated with stalling, not shown to cause it. Report the study against COREQ, a 32-item checklist for interview and focus-group studies (Tong, Sainsbury and Craig, 2007), and apply the anonymization review before anything is shared.
Missing, declined and unsuccessful journeys
| Case | How it is counted |
|---|---|
| Every seat declined | In the frame; furthest stage from the team's own dated records; no reasons reported |
| The CEO declined a company-side interview | Normal; the journey uses the referral side only |
| A seat withdrew after the interview | Their notes are deleted; the journey stays in the counts |
| The owner declined to ask the CEO at all | A journey that ended at stage one, and often the most useful interview |
| Still open on the cut-off date | Held until it closes or reaches 120 days after the ask |
Stop rules. Pause and rewrite prompt 4 if two interviews in a row need a strike for record detail. If a document is kept by mistake, delete it, date the incident and report it in the methods. If more than half of all seats decline, publish stage counts only, because reasons from a minority could mislead.
Decision rule, set in advance. An avoidable code that appears in at least a third of journeys, and in at least three, earns a separate experiment on a fix. Fewer than that, and it is recorded and watched. Results come from one team in one period; nothing here predicts another network's friction.
One volunteered detail, struck (Illustrative)
Illustrative. A generic journey with invented weeks.
The relationship owner said: "It died in week six because their counsel found that most of the service tickets name the customers, and the CEO didn't want the fight." The note reads: "Week 6: the company's counsel raised a question about whose records they were; the CEO stopped. Next step held by the company. Speaker's words on the records struck." Coded R1, the friction map's code for doubt about whose records they are, which the map treats as protective; the journey ended as stopped. The study learns that the stop was the company's own sound call, and learns it without holding a word about the tickets.
What stays with the company and with SourceX
The study never answers the questions that decide an introduction. The records are the company's, and its CEO decides whether to explore licensing them. SourceX's published screen asks for a company with 50 or more full-time employees at its peak, contractors excluded, years of its own documented records and the right to license them. 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 records match current demand and what any license would cover are settled in SourceX's review. Section 7 of the terms adds that partners never receive other partners' records, applicant documents or confidential buyer information, so the study cannot draw on them either.
Limits of looking back
- Memory fades and tidies. Retrospective accounts smooth out waits; the dated-records check catches some of that, not all.
- Who agrees to talk. Owners proud of a journey may say yes more readily than owners whose ask went badly.
- One side of the story. Without company-side interviews, friction inside the company is seen only as a wait.
- Striking costs detail. Some explanations become vaguer once record facts are removed. That trade is deliberate.
- Small numbers. Two quarters of one team's asks may yield a few dozen journeys; treat every count as a description.
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 the study points to a company whose CEO wants the conversation, the partner who will submit should register, recheck the published screen on who qualifies and read the program terms on credit and duplicates first.
Studying referral friction belongs to the referral side of a fund; judging SourceX as an investment is a separate exercise, and materials for it can be requested through the SourceX investor overview.
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.