This is for the growth lead at a venture platform who is about to invite operating advisors, scouts, alumni founders and LPs to register as SourceX partners, and who suspects that leading with the reward will fill the pipeline with companies that do not fit and CEOs who did not want the call.
Why asking people their motives cannot settle it
The obvious study asks each new partner why they joined and later compares the introductions of those who chose "the reward" with everyone else. It fails because the two groups differ in more than motive: tenure at the fund, how many CEOs they know loosely, how much spare time they have.
Economists met the same knot in pay design. Lazear's study of Safelite tested the predictions that switching to piece rates would raise average output and also draw a more able workforce (our paraphrase of the abstract): a pay scheme changes who shows up as well as how hard they work. A referral program has the same two channels. A reward-first message can change who registers, and it can change which CEOs registrants choose to approach. Randomizing the invitation shows the combined effect of those channels on fit, free of the differences between people. It does not split the two channels apart, and this page does not pretend it can.
There is also a specific reason to worry. In four experiments reported in the Journal of Marketing, Ryu and Feick found that rewards raised the likelihood of a referral, and did so most for weak ties and weaker brands (our paraphrase of the abstract). For a data introduction, a weak tie is a CEO the introducer barely knows: the person whose headcount, records and appetite the introducer is least able to judge. That turns the worry into a prediction you can check: if reward interest pulls introductions away from fit, the reward-first arm will show more weak-tie introductions and a smaller share that qualify.
Two invitations that differ only in where the reward sits
Both invitations carry the same five blocks, worded identically. Only their order changes.
| Block | Reward-first invitation | Fit-first invitation |
|---|---|---|
| The reward: by default the greater of $50,000 plus 5% and 15% of what SourceX actually pays the introduced company, with no cap, paid separately from SourceX's own share and never taken from the company's payout; a lead, meeting or signature alone triggers nothing | Subject line and first paragraph | Third paragraph |
| Who decides: the CEO decides whether the company talks to SourceX; SourceX decides fit after its review | Second paragraph | First paragraph |
| What a company needs: 50+ full-time employees at peak, contractors excluded; records it created over years of operation; rights to license them; the location rule | Third paragraph | Second paragraph |
| How credit works: the Channel Partner Agreement is signed before any submission, and the first valid introduction inside 90 days is credited when the company files a verified application, judged from dated records rather than clicks | Fourth paragraph | Fourth paragraph |
| What nobody does: collect files, exports or logins for SourceX | Fifth paragraph | Fifth paragraph |
The location rule appears in both, word for word: 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.
Hold everything else fixed: sender, send day, length within ten percent, links and the subject line's length. A separate test of how big the reward looks, rather than where it sits, is the calculator anchoring study; a test of opening angles in messages to CEOs, rather than to future partners, is the two-angle comparison.
The motivation-versus-qualification study
| Element | Specification |
|---|---|
| Decision | Which invitation goes to the rest of the network |
| Hypotheses, written before sending | The reward-first arm registers more people; qualified introductions per person invited do not differ; the reward-first arm's introductions include more weak ties and a smaller share that qualify |
| Unit | One network member invited to register, never a company |
| Assignment | A logged random split inside each role list: operating advisors, scouts, alumni founders, LPs |
| Window | Introductions submitted on days 1 to 120 after the invitation |
| Read date | Day 210, so the last introduction in the window has its full 90-day attribution window |
| Primary measure | Qualified introductions per 100 people invited: verified applications credited to an invitee and not closed at the read date |
| Secondary measures | Registrations, and introductions submitted, per 100 invited; the share of submissions that qualify; the weak-tie share; days to a first introduction |
| Analysis | By assigned arm, whether or not the person registered |
The introduction log, one row per submission, records the arm; the introducer's own rating of the tie (works with the CEO at least monthly, has met them a few times, or reached them through someone else); how many of the four fit facts were known at submission (size at peak, own records, rights, the CEO's written yes), drawn from public information and what the CEO said, never from records; the date of the CEO's yes; and the dashboard stage on the read date.
The motive question, optional and asked once after registration on the platform's own form: "Which of these mattered most in your decision to register? The possible reward; helping a company I know; my role at the fund; curiosity about AI data; something else." Tabulate it inside each arm to see whether the reward-first invitation shifted the mix of stated reasons. Never use it to rate a person, never send it to SourceX, and delete it with the log when the study closes.
Whoever registers also clears outside pay with their own firm before introducing anyone; the operating partner conflicts page lists who to ask by role.
The founder side: did the introduction feel like it was about the company?
Fit is half of the cost of reward interest. The other half lands on the CEO. In a study of referral reward programs, Wirtz and colleagues found that metaperception suffers when an incentive is present and improves with the tie strength between the person giving a referral and the person receiving it (our paraphrase of the abstract). If anything similar holds for data introductions, a reward can cost the introducer standing with the CEO, so the CEO's side is measured too.
A week after each CEO's yes, the introducer sends two optional questions to a form the platform sees only in aggregate:
- Did the person who introduced you explain who decides and that they could be paid? (yes, no, not sure)
- Did the introduction seem to be about your company's situation? (yes, partly, no)
Each introducer also answers one item at submission: how comfortable were you raising this with the CEO, from 1 to 5.
Four guard events are counted in both arms with a target of zero: an introduction that reached SourceX before the CEO agreed, a CEO complaint about pressure, anyone offering company records, and a CEO who believed the introducer could secure acceptance. Duplicates, companies SourceX already knew, are counted too; a jump in one arm means people reached for familiar names.
Reading the two denominators together
This is the step that separates reward interest from fit. Read per person invited, and per submission, side by side.
A worked reading (Illustrative). Invented counts that show how to read the table; no results exist.
| Count at day 210 | Reward-first arm | Fit-first arm |
|---|---|---|
| People invited | 150 | 150 |
| Registered | 30 | 18 |
| Introductions submitted | 12 | 6 |
| Qualified introductions | 4 | 4 |
| Weak-tie introductions | 7 | 1 |
Per person invited, the arms tie. Per submission, a third of the reward-first introductions qualified and two thirds of the fit-first ones did. The reward-first invitation added six introductions, six more weak ties and no additional qualified introduction. That is reward interest without fit: more asks of CEOs who barely know the introducer, more work in SourceX's review, nothing gained. The per-person count alone would hide it, and the per-submission count alone would make a shorter pile look like a win; the screening cohort analysis explains why the denominator must stay fixed.
| Pattern at the read date | Reading | Decision |
|---|---|---|
| Reward-first ahead per person invited; weak-tie share and CEO answers similar | The reward's position helps people act without pulling them off fit | Send the reward-first invitation and keep the CEO check running |
| Tie per person invited; reward-first has more submissions, fewer qualifying and more weak ties | Reward interest is adding introductions that do not fit | Send the fit-first invitation |
| Fit-first ahead per person invited | Leading with the criteria helps people choose well | Send the fit-first invitation |
| No difference anywhere | Position is not what drives behavior | Keep whichever produces fewer misunderstandings in CEO answers |
| Any guard event | A wording problem | Fix it before reading anything else |
Floors and stop rules
- Publish rates only when each arm holds at least 100 people invited; below that, publish counts.
- Do not read the share that qualifies until each arm has at least 15 submissions; with fewer, a single introduction moves the share by about seven points.
- If both arms together hold fewer than 10 qualified introductions at the read date, report every count and the verdict "no answer yet".
- Each person gets one invitation. Reminders, if any, are identical in both arms.
- Pause both arms if an introduction reaches SourceX before its CEO agreed; stop if anyone collects company records; pause an arm after a CEO complaint about pressure; start a new run if SourceX publishes new terms.
- Never stop early because the reward-first arm's registrations look strong.
Where this experiment can mislead
- Small nudges. Moving a paragraph is a light touch, and a network of a few hundred may not show its effect.
- Combined effects. The design shows what the invitation does overall; it cannot say how much came from who registered and how much from how they behaved.
- Self-reports. Few people tick "the reward" even when it mattered, and tie strength is the introducer's own rating.
- Fatigue. Members invited to many programs may ignore both versions, shrinking any difference.
- Review sits with SourceX. A qualified introduction is a verified application still open, not an accepted company or a payable reward, and SourceX's programs and terms can change during the window.
Methods note. The design rests on SourceX's terms (version 2026-10-22.2, since replaced by version 2026-11-04) and screening page, read on 2026-10-28, and on the three studies cited above for its predictions. It measures registrations, introductions, dashboard stages, tie ratings and optional survey answers. It does not measure demand for any company's records, acceptance, or any payout, and no figures are published because none have been collected.
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.
Before either invitation goes out, register and sign the Channel Partner Agreement yourself, copy the fit block from who qualifies and the reward block from the program terms so both arms carry the same dated wording; other measurement designs are in the guides hub.
Testing invitations is a referral decision. Anyone assessing SourceX as a possible investment has a separate route: the SourceX investor overview, where materials can be requested and the company says it shares company and commercial metrics directly.
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.