This page is written for a research director at a venture platform or operating network who suspects that partners go quiet while they wait for news. Neighboring pages answer different questions. How many introducers come back, cohort by cohort, is counted in the partner cohort retention report, which also sets out which second appearances count as a return. Whether later introductions are better than a partner's first is the subject of the learning-curve plan. Keeping one partner's own records straight belongs to the repeat partner activity worksheet. This study asks one thing those cannot: does the news a partner has, or lacks, line up with whether and when they introduce again?
Why a retention count cannot answer the feedback question
A retention count tells you who came back. It cannot tell you what the people who came back knew, and in this program they often knew very little. A reward becomes payable only after the Acquisition Partner's acceptance, cleared payment to SourceX, payment to the company and finance approval, under the deadline in section 5 of the terms. Long before any of that, a partner has usually decided whether to raise SourceX with a second CEO.
That timing creates a trap. If the analysis sorts partners by how their first company finally turned out, it credits a partner's second introduction to an outcome that may not have existed when they made it. The study avoids the trap by recording the partner's view as it stood on each date, and comparing partners only on that.
What the partner can see is limited by design. SourceX's referral page says the dashboard shows each referral's stage and a reward status (Estimated, Eligible, Approved, Paid), with Paid shown only after an actual payout is recorded, and section 4.6 of the sample agreement says the partner receives what is reasonably needed to track its own referral, in generic stages or reference codes. The study works with exactly that view, and nothing richer.
What counts as a return
A return is a later introduction of a different company, classified after SourceX's duplicate review as new or not. A partner rarely knows at the time whether SourceX already knew a company; section 3 of the terms says "A relationship includes any earlier lead, account or email correspondence between SourceX and the company, however it began." Repeat licenses and renewals by a company already introduced are recorded as context only: section 4 keeps them under one cap per company, which "does not reset for a new agreement," and SourceX's process page says returning suppliers reuse their Master agreement and sign a new Dataset Deal Confirmation. A sister company in the same group sits in its own category, because one cap covers a parent and its subsidiaries and how earlier contact elsewhere in the group affects credit is settled by SourceX in duplicate review, not by this study.
The partner behavior cohort protocol
Entry. A partner enters on the date of their first introduction, submitted through the dashboard or recorded in writing, during a two-quarter entry window.
Follow-up. Twenty-four months from entry.
Time at risk. Only days under a signed Channel Partner Agreement. This is a study rule grounded in the documents: section 1 of the terms asks for a signed agreement before any introduction is submitted, section 9 says the agreement lasts a year and renews only with fresh acceptance and signatures, and section 5.2 of the sample agreement says a referral arriving after expiration needs renewal or a new agreement to be protected under its section 2. Days between expiry and renewal are removed from the denominator and reported on their own line.
The event. Each later introduction of a different company, with two counts side by side: all of them, and those SourceX treated as new.
The exposure, called the feedback state. For each partner on each day at risk, the furthest point reached by any earlier introduction, as the partner could see it.
| Feedback state | What the partner could see | Read from |
|---|---|---|
| F0, no word | No stage change since the introduction | Dashboard history |
| F1, moving | A stage change, no reward status yet | Dashboard history |
| F2, estimated | A reward status of Estimated | Dashboard history |
| F3, eligible or approved | A reward status of Eligible or Approved | Dashboard history |
| F4, paid | A reward status of Paid | Dashboard history |
| FX, closed | A stage showing the referral has ended, where the dashboard shows one | Dashboard history |
The state changes only on a dated dashboard event, so it can be reconstructed for any day after the fact.
| Other variable | Definition | Source |
|---|---|---|
| Partner type | Investor, adviser, association or other, self-described | Entry survey |
| Network band | CEOs the partner speaks to in a year, in bands | Entry survey |
| First company's route | Whether the CEO asked for the introduction or the partner raised it | Entry survey |
| Dormant spell | Six months at risk with no introduction | Derived |
Sampling frame and consent
The frame is every partner whose first introduction falls in the entry window. Each is asked to consent in writing to the use of their records and to a three-question entry survey; partners who decline appear only as a count, so readers can see how far the cohort differs from the frame. Companies are never named: introductions carry random codes, an anonymization review runs before any release, and cells with fewer than five partners are suppressed.
Analysis: landmarks first, recurrent-event models second
Landmark comparisons. At months 3, 6, 9 and 12, take every partner still at risk who has not yet made a later introduction, group them by their feedback state on that day, and follow each group for the next six months at risk. A partner's group is fixed at the landmark and never revised with later news, which is what keeps the comparison honest.
Recurrent-event models. For the full 24 months, the plan follows the family of methods a tutorial in the International Journal of Epidemiology sets out. In paraphrase of its abstract, Amorim and Cai cover conditional models (Andersen-Gill and the Prentice-Williams-Peterson total-time and gap-time models), marginal means and rates models, frailty models and multi-state models. The primary model is a gap-time model for the wait between introductions, with the feedback state as a time-varying covariate and a frailty term because partners plainly differ; a marginal rate of later introductions per partner-year at risk is reported beside it. Both are fixed before the window opens.
Every estimate is an association. Partners with close CEO relationships may both hear news sooner and introduce more often, and nothing in the design separates the two.
The reporting template, empty until month 24
| Landmark | Feedback state at landmark | Partners still waiting | Introduced again within six months | Of which new to SourceX | Lapsed days removed |
|---|---|---|---|---|---|
| Month 3 | F0 to FX, one row each | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| Month 6 | F0 to FX, one row each | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| Month 9 | F0 to FX, one row each | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| Month 12 | F0 to FX, one row each | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
Nothing is published before month 24. The full table then goes out with zeros, the frame count, the consent rate, the model estimates with their intervals, and the versions of SourceX's terms in force during follow-up, since section 11 allows new ones. Existing SourceX results are cited only under the standard in partner results research.
Counterexamples the report must show
- Partners who never introduce again, whatever their feedback state.
- Partners who introduced again within days, before any news could exist.
- Partners whose only later introductions were duplicates.
- Partners who came back right after a closing stage on their first company.
- Partners whose return followed a renewal rather than any dashboard event.
Results that would change how repeat activity is read
- No difference across feedback states. Then news about earlier companies is not what brings partners back, and investing in faster status updates rests on a guess.
- Returns cluster just after renewals. Then the one-year agreement, not feedback, sets the rhythm.
- F0 partners return as often as F4 partners. Then partners act on their own read of a company's fit, not on outcomes.
- Most later introductions are duplicates. Then repeat activity mostly reaches companies SourceX already knew.
Risks and limits a fund should weigh
- Consent bias. Partners who agree to be studied may be the most engaged.
- Sparse late states. Few earlier introductions may reach Paid within 24 months, so F4 rows may stay too small to report.
- Changing labels. If the dashboard's stages or statuses are renamed during follow-up, the mapping to F0 to FX is dated and published.
- Self-described partner type. Investors who also advise companies may choose either label.
- Status is not value. A dashboard status says where a referral stands, not what it will become, and says nothing about SourceX's business.
What a partner can keep in the meantime
One dated line per company with your own last step kept apart from what SourceX shows, as the activity worksheet sets out, plus the day your agreement was signed and the day it renews. If earlier introductions went quiet, the custody-change study shows how to record who held each next step. Before a next introduction, the company should plausibly fit: a full-time headcount of 50 or more at its peak, contractors excluded, years of records it created itself, the freedom to license them, and a CEO who has asked to hear more. 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.
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 your second company is already in view, register or renew your agreement before anything is submitted, have the CEO compare the company with who qualifies and check the program terms on duplicates; more study designs are in the guides hub.
Studying how partners return is referral research. Readers evaluating SourceX as a company can request investor materials, a separate matter, 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.