Measure the time from permission to a qualified introduction: a stage-time cohort that treats a withdrawn yes as an ending, not a slow case

Short answer

Start the clock on the day the company's CEO says they want the conversation, and stop it when SourceX credits the company's verified application to you. Every other ending stays visible: a CEO who withdraws, or a company SourceX already knew, closes the row as an ending rather than counting as a slow case. For one quarter's cohort, report how many companies were credited by day 30, 60 and 90, and print the denominator, the closed rows and the open rows beside each figure.

Visual briefFull-size diagram
Twelve Illustrative company lanes measured in days from the CEO's permission, with lines at day 30, 60 and 90: 3 of 12 credited by day 30, 5 of 10 by day 60 and 6 of 10 by day 90, three lanes closed without credit and three still open.
Illustrative cohort of twelve generic companies; the counting rules follow this page and no real referrals or outcomes are shown.

Credit is the last event a fund's own work controls. SourceX's review of the company follows it and runs on a separate clock, which the stage duration benchmark measures with its own cohort rules. This page times the stretch before that, where the waiting belongs to the fund and the company rather than to SourceX.

Five dated events between a CEO's yes and a credited application

A stage is the gap between two events that each carry a date and a piece of evidence. Define them once and write the definitions above the report.

EventWhat must be trueDated evidenceWhose time it measures
Ask approved (before day 0)The relationship owner agrees to raise the idea with the CEOA note in the fund's logThe fund's, before the clock
Permission (day 0)The CEO wants the conversation and knows SourceX will be introducedThe relationship owner's written note of the CEO's answerStarts the clock
Link sentThe partner sends the referral link or introduction with the CEO's knowledgeThe sent messageThe fund's
Application submittedThe company applies itself, as SourceX's process requiresThe company's confirmation, or the referral appearing in the portalThe company's
CreditedSourceX verifies the application and credits it to you as the first valid introductionThe referral shown as yours in the portalSourceX verification; stops the clock

Three things can end a row without credit. The CEO or the relationship owner withdraws after permission. Duplicate review finds that SourceX already had an application, a relationship or an earlier valid introduction for the company, and the program terms count any earlier lead, account or email correspondence as a relationship. Or the cutoff arrives first, and the row is simply still open.

Population. Every company whose permission date falls inside one calendar quarter, one row per company. Reporting period. Read every row on one cutoff date, 30 days after the quarter ends, and never add a later permission to a closed cohort. What credit is not. Sending a link or an email does not by itself make an introduction eligible for a reward, and credit is not a reward either: under the terms, a lead, meeting or signed agreement alone does not earn a payable reward. Attribution is decided from dated records, which is one more reason to keep these dates. The qualified referral definitions set credit and payment side by side.

The stage-time cohort analysis, built row by row

The asset has two parts: a roster that shows every company, and a horizon table that counts them without hiding anyone.

Roster: one quarter's permissions (Illustrative). Twelve generic companies with invented dates. All numbers are days from permission.

RowCompany (Illustrative)Link sentAppliedCreditedEnding at cutoffClock stopped or cutoff reached
A180-person claims administrator169Credited9
B240-person freight broker21216Credited16
C95-person software implementation firm11923Credited23
D410-person contact center operator93034Credited34
E130-person engineering consultancy34551Credited51
F300-person property manager148088Credited88
G160-person building-products distributor2NeverNoneWithdrew: the CEO paused12
H220-person insurance agency group417NoneDuplicate: earlier email with SourceX20
I75-person dental billing service5NeverNoneWithdrew: the board asked to wait40
J500-person logistics group2Not yetNot yetOpen31
K140-person HVAC services company21Not yetNot yetOpen47
L260-person manufacturer's rep agency360Not yetOpen: awaiting verification95

Horizon table. For each horizon, the denominator is every company that either resolved by then or was watched for at least that long. A row still open and watched for less time is listed as not yet observable; it is never dropped silently.

Horizon from permissionCompanies observableCreditedClosed without creditStill openNot yet observable
Day 30123270
Day 60105322
Day 90106312

Closed rows stay in the denominator. It is tempting to treat G and I as censored, as if their clocks had merely stopped, but that assumes those two CEOs would have gone on to apply at the pace of everyone still waiting. Austin, Lee and Fine, writing in Circulation, show that the Kaplan-Meier estimate overstates the cumulative incidence of an event when competing events are treated as censored, and recommend cumulative incidence functions instead. Counting endings as endings, as the table does, is the small-cohort version of that advice.

The roster also shows why a single average misleads. The six credited companies took a median of 28.5 days, and a committee that hears only that figure hears "about a month". Yet only 3 of 12 had been credited by day 30, and at the cutoff 6 of 12 had not been credited at all: three closed without credit and three are still open.

Three clocks, three owners. Split the stretch so each delay has a name attached.

ClockWho owns the paceRows measuredMedian daysNamed exception
Permission to linkThe fund123K: link sent on day 21, still open
Link to applicationThe company819.5F: 66 days between link and application
Application to credit or closureSourceX verification74L: open 35 days after applying, left out of the median and named

Only the first clock is fully the fund's. In this invented cohort the three rows that waited a week or more for a link include the slowest credit and one row still open. Twelve rows cannot show that the late link caused anything, but it is the one delay the platform team can remove without asking anyone else.

Reconciling companies, repeats and cases still open

  • One row per company per cohort. If a CEO withdraws and returns months later, close the first row as withdrawn and open a new row in the later cohort marked as returning. Any combined total counts the company once.
  • Duplicates are endings, not verdicts. A row closed by duplicate review says SourceX knew the company first. It says nothing about the company's records.
  • Groups. The cap treats a corporate group as one company, so a parent and its subsidiaries share it. Whether a sister company's earlier contact with SourceX affects credit is not settled in the published terms; SourceX decides it in duplicate review, so record the group link and wait for the portal.
  • Open rows name what they await. J awaits an application, K awaits an application after a late link, L awaits verification. An open row without a named next event is a row nobody owns.
  • Monthly portal check. Every row marked credited in the log must appear as yours in the portal, and every portal referral must have a row with a permission date. The terms also let SourceX communicate directly with a referred company without affecting valid attribution, so a fast application may owe nothing to the fund.

What a CEO does after permission is their call, and the fund records it without pressure; what happens if the owner changes their mind covers the company's side.

What the quarterly note says, and the reward line it leaves out

A useful note reports learning in the cohort's own units. For the roster above:

Q3 cohort, read 30 days after quarter end (Illustrative). Twelve CEOs gave permission. By day 30, 3 of 12 companies were credited; by day 90, 6 of the 10 we could observe that long. Three rows closed without credit (two withdrawals, one duplicate) and three remain open. Our own lag from permission to link had a median of 3 days, but three rows waited a week or more; from next quarter, links go out within two working days of a yes.

Leave money out of this note. Credit happens long before any supplier payment, the terms describe portal and calculator estimates as projections rather than promises, and multiplying a quarter's credits by four and by an assumed payout produces an annual figure no event supports. Rewards belong on the outcome dashboard, entered only once paid; the dashboard that separates interest from outcomes keeps those columns apart from this one.

Method notes and where this clock can mislead

Methods note. The stages, endings and credit rule are drawn from two SourceX pages read on 2026-10-26, who-qualifies and the program terms in Version 2026-10-22.2, since replaced by version 2026-11-04; the competing-endings rule follows Austin, Lee and Fine (2016). The roster and every count are invented to show the arithmetic. The analysis measures days between dated events and the share of a cohort credited by fixed horizons. It does not measure SourceX's review after credit, demand for any company's records, what a license would pay or whether any reward follows, and SourceX has not published stage-time or conversion figures for this program.

Where it can mislead:

  • Small numbers. One row moves a twelve-company cohort by eight percentage points. Print counts, not just percentages.
  • Self-recorded starts. The permission date comes from the relationship owner's note. A note written a week late shortens every clock; require the date of the CEO's own message.
  • Speed pressure. A target on days-to-credit can push a CEO toward an application they were not ready for. The CEO decides when the company applies, and a withdrawal is a legitimate ending.
  • Unequal follow-up. Compare cohorts only at the same horizon. A quarter read at day 30 always looks slower than one read at day 90.
  • Changing definitions. The terms are versioned. Note the version in force for each cohort.

Which companies belong on the clock at all

A row starts only when a company has passed the published fit facts as far as the fund can tell, and its CEO, the person who authorizes a licensing discussion and signs for the company, has said yes. Those fit facts are a peak full-time headcount of 50 or more with contractors set aside, documentation of the company's operating history, rights to license the records, and an authorized contact. 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.

Everything the fund learns before permission stays at metadata level: the record types, a rough volume, the span of years, the system and who administers it. Nobody asks for files, samples or logins. Whether the records match a current program, whether contracts limit their reuse and what a license would cover remain unknown until SourceX reviews the company. The person who submits must already have signed the SourceX Channel Partner Agreement; a signed copy that still bears the earlier title, Channel Partner Non-Circumvention and Non-Disclosure Agreement, remains in force as it was signed. For preparing the introduction itself, see how to prepare a qualified company introduction; for the portfolio side of the program, the overview for venture platform teams; and for related measurement pages, the guides hub.

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 CEO in your network has said yes and the fit facts you can check all pass, register for a partner link, run the company through who qualifies and read the program terms, then send the link within days rather than weeks so your own clock stays short.

Timing a referral cohort is platform operations work; whether SourceX itself deserves a look as an investment is a different question, and for that one, investor materials 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.

  1. Step 1Share your linkSend your personal link to a company you know.
  2. Step 2Company appliesThe company applies itself at /apply.
  3. Step 3Company signs and is paidThe company signs, delivers its data and is paid by SourceX under its agreement.
  4. Step 4You get your rewardthe greater of $50,000 plus 5% and 15% of the actual supplier payout, with no cap. A reward becomes payable after SourceX has paid the referred company under its signed agreement and the payout conditions in the program terms are met, including finance approval and your tax and payment details.

Common questions

Why does the clock start at permission rather than at the referral link?

Because the CEO's yes is the first event the company controls, and the gap between that yes and the link is the fund's own delay. Starting at the link hides exactly the lag a platform team can fix.

Is a credited application the same as a reward?

No. Credit settles which partner, if any, the company is attributed to. A reward exists only after SourceX has paid the company and the payment conditions in the terms are met, which can be months later or never.

How many companies does a cohort need before medians mean anything?

No threshold has been published for this program, and a dozen rows cannot support percentiles. Report the counts by horizon with their denominators and list the rows; add medians once most of a cohort has resolved.

Sources

  1. refer.sourcex.si/terms
  2. refer.sourcex.si/who-qualifies
  3. ahajournals.org/doi/10.1161/circulationaha.115.017719
  4. sourcex.si/investors

Free resources

Know a company with valuable proprietary data?

Become a referral partner, get your link and introduce the company’s CEO (in an owner-run company, usually the owner). Referral partners can join from most countries. We can't work with partners in countries or regions under U.S. sanctions or on U.S. restricted-party lists. The companies you introduce should be based in the U.S., Canada and the EU, or in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia with at least 70% of their data in English.

Refer a company →

I own a business

Explore licensing your company's data directly to SourceX for AI training and evaluation. Start a short assessment; no uploads needed.

Start an assessment