The scarce input is the platform team's attention, not the reward. A dedicated owner is a standing claim on it, so decide as you would any headcount: on evidence from a small pilot, revisited on a date, apart from any hope about what the program might pay.
What the founder gets, and what the team pays in attention
The founder's benefit is concrete and limited: a structured way to learn whether the company's own records could be licensed directly to SourceX for AI training and evaluation. SourceX reviews history, documentation, rights and relevance; the company controls whether to proceed and on what terms; the platform team never exports, uploads or describes records. Fit is narrow, and SourceX judges it, not the platform team: a company needs to have reached 50 or more full-time employees at its peak, not counting contractors, plus years of records it created, the rights to license them, a CEO who authorizes the discussion and a match with a current program. Funding status appears in none of those checks.
The team's cost is time, and even large platform organizations divide that time by function. Andreessen Horowitz says it pioneered the platform model, with what it describes as the largest team of operators in venture, spanning marketing, talent, legal and policy. A new workstream in a team organized that way needs a home, or it becomes everyone's side task and nobody's job.
Ownership is where partnership work tends to break. When PartnerStack and Wynter surveyed 100 senior leaders at B2B SaaS companies for their 2026 report, 69% planned to increase partnership investment, yet alignment between teams was the most-cited blocker to partner-driven revenue, at 37%, ahead of having no clear partner program, at 20%. Those are software companies selling through partners, not venture platforms, but the lesson travels: partnership work usually fails at the seams between people, not for lack of a program.
The compensation is too uncertain to staff against. Even the upside is late and conditional: by default a company's first qualifying transaction pays the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap, with the USD 50,000 floor applying once per referred company and repeats within 12 months paying 15%, and none of it is payable until the Acquisition Partner has accepted the dataset, SourceX holds cleared payment, the company has been paid, finance has approved and the partner's tax and payment details are on file. A lead, meeting or signature alone earns nothing. Every model on this page therefore books referral income at $0.
Three ways to own the work
| Ownership level | Who | What they do | Fits when | Breaks when |
|---|---|---|---|---|
| Occasional task | Whoever hears of a likely fit | Checks the published criteria, asks the relationship holder, sends one introduction | One or two conversations a quarter | Founder questions wait for days; two colleagues approach one CEO |
| Named part-time owner | One platform team member with a stated share of time | Keeps one company list and time log, coordinates asks, submits introductions, follows up with founders | Several conversations open at once | The owner's other programs crowd it out at peak weeks |
| Dedicated owner | A role whose main job is this workstream | Adds network-wide screening beyond the portfolio, LP and co-investor paths, and quarterly reporting | Two quarters of sustained volume and useful founder feedback | Volume falls and the role starts generating asks to justify itself |
The middle row is where most teams should land first. It costs a share of one person's time and fixes the two failures the first row produces: questions with no owner, and duplicate approaches. SourceX credits one partner per company, so a single list with a single owner also settles who submits each introduction before anyone writes to a CEO.
The ownership threshold decision guide
Five signals, each taken from a log the owner keeps. Move up a level only when the signals in that column hold for the whole review period; drop to "stop" when the stop column applies.
| Signal (logged) | Stay occasional | Name a part-time owner | Consider a dedicated owner | Stop |
|---|---|---|---|---|
| CEO or founder conversations open at one time | 0 to 2 | 3 to 8 | 9 or more, for two quarters | None after the pilot's planned asks |
| Logged hours per month | Under 8 | 8 to 30 | Over 30, for two quarters | Hours logged with no CEO conversation to show |
| Founder questions waiting more than 5 business days | None | One or more a month | Recurring, even with a part-time owner | Not applicable |
| Duplicate approaches to one CEO | None | Any | Repeated across several relationship holders | Not applicable |
| Founders answering yes to "Was this worth your time?" | Most | Most | Most, and some ask for more | Most say no |
Method note. We drafted the three levels and five signals on 2026-10-26, working from the program terms (Version 2026-10-22.2, since replaced by version 2026-11-04), the screening guidance on the who-qualifies page and the how-it-works page, and took the cost of unclear ownership from the PartnerStack and Wynter survey. The numeric thresholds are Illustrative starting values, not benchmarks: SourceX has not published conversion rates, typical hours per introduction or how often introductions lead to licensing deals, and no comparable platform-team data has been published. Replace each threshold with the readings from your own pilot. Nothing here measures demand for any company's records, which SourceX assesses in review.
Work at each referral stage, with income at $0
Hours per company, by stage. The figures are Illustrative assumptions to replace with logged time; the stage list follows the published process, in which the company applies itself and SourceX runs the review.
| Stage | Platform team task | Illustrative hours per company |
|---|---|---|
| 1. Screen | Check size at peak, history, rights and the CEO path against the published criteria | 0.5 |
| 2. Ask the relationship holder | Agree who asks the CEO, and whether to ask at all | 0.5 |
| 3. CEO conversation | Explain the program; the CEO decides whether to talk to SourceX | 1.5 |
| 4. Introduce | Written introduction and portal submission | 0.5 |
| 5. During review | Route the founder's questions to SourceX; handle no records | 1.0 |
| 6. Track and close | Read the dashboard stage and tell the founder where things stand | 0.5 |
The zero-revenue pilot base case (Illustrative). In one quarter the team screens 20 companies (10 hours), asks relationship holders about 8 (4 hours), holds 5 CEO conversations (7.5 hours), introduces 3 (1.5 hours), supports those 3 through review (3 hours) and tracks them (1.5 hours). That is 27.5 hours, about 9 a month, with $0 of referral income booked. On the guide, about 9 hours a month and 3 to 5 conversations open at once point to a named part-time owner, not a dedicated one. A deeper hours model, with readings for expanding or pausing the whole program, is in whether data referral work is worth a platform team's time.
Two pilots: one earns an owner, one should stop (Illustrative)
Illustrative. Generic funds, invented counts.
Fund A screens 20 companies across its portfolio and its LPs' holdings, as in the base case. Five CEOs take the conversation, four say it was worth their time, and three introductions go in. Logged time averages 9 hours a month. Decision: name a part-time owner for the next two quarters, set response commitments for founders (the lightweight service level is one template), and reconsider a dedicated role only if the next two quarters cross that column.
Fund B screens 30 portfolio companies. Twenty-seven have fewer than 50 full-time employees at peak, two keep most of their records inside customer tenants, and the one company that passes every check belongs to a founder whose board member declines to raise it this quarter. After six weeks: no CEO conversation and 6 logged hours. Decision: stop, and assign no owner. Record each failed check with the trigger that would change it, such as full-time headcount passing 50, and screen again when the published profile changes.
Fund B is not a failure of effort. It is the pilot doing its job: an inexpensive way to learn that this fund's network does not yet hold the companies the program screens for.
Expand, pause and stop, written before the first ask
- Expand to a dedicated owner after two consecutive quarters above the part-time column on conversations and hours, with founders mostly saying the help was worth their time and no introduction made without a CEO's knowledge.
- Pause when SourceX's program criteria change (criteria follow each program and can move), or when the owner's other programs peak, with a restart date written down.
- Stop when the pilot's planned asks produce no CEO conversation, when founders mostly say it was not useful, or when an introduction goes out without the CEO's knowledge.
- Never set an owner's target in rewards. Payment follows acceptance, cleared payment and finance approval that can arrive long after the introduction, and a reward target turns the owner into a salesperson for SourceX.
What could make these thresholds wrong
- Demand moves. SourceX checks relevance to current data needs and applies each program's own criteria; a well-run pilot can find fewer fits next quarter for reasons outside the fund.
- Rights. Records that belong to clients are usually not a fit, and that is decided in SourceX's review, not by the owner's screening.
- Long cycles. Rewards are payable only after acceptance, cleared payment, supplier payment and finance approval, so a dedicated role judged on rewards would look idle for months even when it works.
- Self-reported time. Time logs undercount short interruptions; ask the owner to log weekly, not from memory at quarter end.
- Relationship cost. Asking a founder or LP to raise this spends goodwill, and no data on how they react has been published. Record it in the founder question.
A team that decides the work deserves an owner can hand that person the overview for venture platform teams first, then the first-month plan and the guide to how platform teams explain data partnerships. The trade-off against programs the team already runs is scored in the opportunity cost of another founder program.
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.
Run the pilot before naming anyone: register for a partner link under the person who will keep the list, screen each company against who qualifies, read the program terms, and let the logged signals decide the ownership level at the end of the quarter. More platform guides are in the guides library.
Staffing referral work is an operating choice for the platform, while putting money into SourceX would be an unrelated investment decision; anyone weighing that can ask for investor materials 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.