How to recognize a partner network that is starting to compound: six cohort signals, and the growth that only looks like it

Short answer

A partner network is compounding when each round of introductions makes the next one better: the same introducers come back with another company that fits, the share of introductions reaching a verified application rises from one cohort to the next, and screening questions get sharper because past declines taught something. More names, more links and more applications are growth, not compounding. Track repeat introducers and fit rates by cohort. SourceX has not published network-level figures, so this page gives the dashboard and its method rather than results.

Visual briefFull-size diagram
Two Illustrative networks over three quarters: Network A's introductions rise from 6 to 15 to 30 while its fit rate falls from 33% to 27% to 23%; Network B's rise from 6 to 8 to 10 while its fit rate climbs from 33% to 50% to 60%.
Illustrative. Two fictional networks; counts and rates match the cohort table on this page and are not SourceX data.

This is written for whoever runs a group of introducers: a firm whose consultants refer clients, an association, a peer group or an informal circle of advisers who share a partner link. The decision is whether to keep adding members or to invest in the ones you have.

What the program fixes, and what this page proposes

StatementStandingBasis
Credit for a company goes to a single partner, whoever introduced it validly first, and it lands only when the company's verified application arrives inside the 90-day attribution window; later duplicates are not creditedProgram factSourceX terms, checked 2026-10-26
A lead, meeting or signature alone does not trigger payment, and the payment deadline is 60 calendar days after the later of the Acquisition Partner's final written acceptance of the complete, validated dataset and completed handoffProgram factSourceX terms
Whether any introduced company completes a handoff and any reward is paidConditionalSourceX review and a signed agreement
Compounding shows first in leading signals (repeat introducers, fit rate by cohort, avoidable declines), long before rewardsProposed methodEditorial; no network-level figures have been published
TermMeaning here
CohortIntroducers grouped by the quarter of their first introduction
Useful introductionOne where the CEO knew about and wanted the conversation, and the company submitted a verified application
Repeat introducerSomeone who makes a second useful introduction within four quarters of their first
Fit rateUseful introductions divided by all introductions submitted, read at the same age for every cohort
Avoidable declineA no-fit reason the introducer could have checked at metadata level: headcount, record ownership or the CEO's knowledge
ConcentrationShare of useful introductions made by the two most active introducers

What compounds: learning, trust or qualified introductions?

Three things can, and each has a counterfeit that looks the same on a count of activity.

What compoundsHow it shows upSignal to measureCounterfeit
LearningScreening questions get sharper because past declines taught the network somethingAvoidable declines fall from cohort to cohortA longer checklist nobody uses
TrustIntroducers come back; a CEO who had a useful conversation mentions it to peersRepeat introducer share rises; introductions arrive from a previously introduced CEO's circleOne-off favors to a single partner that never repeat
Qualified introductionsEach round produces more companies that pass screeningFit rate rises at equal cohort ageMore introductions at a flat or falling fit rate

Investment in partner channels is rising in other markets: in PartnerStack and Wynter's 2026 survey of 100 senior leaders at B2B SaaS companies, 69% said they are increasing investment in partnerships. Investment is an input. Compounding is about what each unit of effort returns, and adding members is the easiest way to raise inputs without raising returns.

In this program the introducer's job is narrow. The partner makes one introduction; SourceX qualifies and prices the company, licenses the data and pays. So the only things a network can compound are the ones it controls: which companies it chooses, how well it screens them and whether the CEO wanted the conversation.

Track repeat partner cohorts and company-level outcomes

Log every introduction at company level, assign it to the introducer's cohort, and read outcomes at the same age for every cohort.

The repeat partner worksheet already tracks companies, status and rewards company by company (there is no cap). A compounding view adds five columns:

  1. Cohort of the introducer, by the quarter of their first introduction.
  2. CEO knew and wanted the conversation, yes or no, recorded before submission.
  3. Fit facts known before submission: headcount at peak, years of own records, ownership and location, each marked known or unknown.
  4. Outcome: verified application (with date), declined with the stated reason, withdrawn, open, or completed handoff.
  5. Duplicate flag, for companies another partner introduced first.

Read every cohort at the same age, for example two quarters after its first introduction. A cohort four quarters old has had more time for cases to close than one two quarters old, and comparing them flatters the older group. Open cases are not failures; they are unknown. The conversion benchmark design shows how to report them as a range rather than drop them, and the rejection reasons study shows how to classify declines.

Is screening improving as the network grows?

Compare avoidable declines and fit facts known before submission across cohorts read at the same age. The sharpest test is not whether founding members improve. It is whether newcomers start better than founding members did.

A network compounds when its newest cohort starts where its oldest cohort ended.

Illustrative case: one decline that taught the network. A network of eight fractional COOs. Generic companies, invented facts, no outcome implied.

  • Quarter 1. A member introduces a 180-person HVAC services group. The CEO wanted the conversation, but its service history lives in a facilities-management client's portal under that client's contract. SourceX declines: the records belong to the client. The network adds one question to its screen: "Whose system holds the service records, and whose contract governs them?"
  • Quarter 2. A member who joined that quarter uses the question with a 240-person specialty contractor. Its 12 years of job-cost and change-order records sit in its own system under its own contracts. The CEO wants a conversation and the company submits a verified application. What happens next depends on SourceX's review; the network records the case as useful, not as a deal.
  • Quarter 3. Another member submits five companies from a trade-show list in one week. None of their CEOs knew. SourceX's screening guidance says a company is usually not a fit when its decision maker never heard about the introduction, and all five end that way. The network's count of introductions jumps; its fit rate drops.

Quarters 1 and 2 are compounding: a decline became a question, and a newcomer used it. Quarter 3 is growth. A dashboard that only counts introductions would score quarter 3 as the best of the three.

Duplication, concentration and review capacity

Three constraints can make a growing network look stronger than it is, and a fourth makes money the slowest signal you have.

  • Duplication. As members' circles overlap, two of them may know the same company. Only the first valid introduction is credited, and later ones are reviewed as duplicates. A rising duplicate share means the network is saturating its own overlap. Agree who approaches a company before anyone contacts its CEO.
  • Concentration. If two members make most of the useful introductions, the network is two partners and a mailing list. It is fragile, and adding members will not fix it unless the newcomers learn what those two do.
  • Review capacity. Introductions submitted faster than they can be reviewed become open cases, and open cases are not outcomes. The company's own pace matters too: its CEO, counsel and system owners decide how fast a conversation moves. Watch open cases older than 90 days.
  • The reward lag. Rewards arrive last: after the Acquisition Partner's final written acceptance, completed handoff, SourceX's payment to the company, finance approval and the partner's tax and payment details. A later cut or reversal of the supplier payout can lower a reward already recorded, since the reward is recalculated on the reduced payout. Reward totals describe introductions made long ago. Use them as a lagging check, never as the compounding signal.

The repeat-introduction and quality cohort dashboard

Six tiles, each with its denominator and what both readings look like. Read them by cohort at the same age.

TileDefinitionDenominatorCompounding looks likeGrowth-only looks like
1. Fit rate at two quartersUseful introductions divided by introductions submittedEvery introduction in the cohort, duplicates and declines includedRising from cohort to cohortFlat or falling while counts rise
2. Repeat introducer shareIntroducers with a second useful introduction within four quartersEveryone in the cohortRisingFalling as new names join
3. Avoidable decline shareDeclines for headcount, record ownership or the CEO's knowledgeAll declinesFallingSteady
4. Fit facts known before submissionIntroductions with all four fit facts recorded before submittingAll introductionsNewest cohort starts near the oldest cohort's levelEach new cohort starts from scratch
5. Duplicate shareIntroductions reviewed as duplicatesAll introductionsLow and stableRising
6. ConcentrationUseful introductions made by the top two introducersAll useful introductionsFallingRising, or stuck near all of them

Two networks, three quarters (Illustrative). Fictional figures built to show the arithmetic, not SourceX data or a forecast. Both networks start the same way.

QuarterNetwork A introductionsNetwork A usefulNetwork A fit rateNetwork B introductionsNetwork B usefulNetwork B fit rate
Q16233%6233%
Q215427%8450%
Q330723%10660%
Total5113the greater of $50,000 plus 5% and 15%241250%

Network A recruited hard and made 51 introductions for 13 useful ones. Network B grew slowly and made 24 for 12. By the end of the third quarter they had almost the same number of useful introductions, but Network B asked fewer than half as many CEOs to spend time on a conversation, and its next quarter starts from a better screen. Network A is growing. Network B is compounding.

Method note. The tiles and definitions are built from SourceX's published attribution rule and payment conditions (terms, checked 2026-10-26) and its screening guidance (who-qualifies page). They measure a network's own choices and screening. They do not measure demand for any company's records, SourceX's review decisions or payouts, and SourceX has not published network-level cohort figures.

What would show growth without compounding

Any two of these together mean the network is getting bigger, not better:

  • The fit rate is flat or falling at equal cohort age while introductions rise.
  • The newest cohort starts with the same avoidable-decline share as the first one did.
  • Repeat introducer share falls as members join.
  • The top two introducers make most useful introductions, year after year.
  • Open cases older than 90 days grow faster than cases close.
  • Reward totals rise only because of handoffs from much earlier cohorts.

When the recommendation changes: if your network is small, single cases swing every percentage, so keep the case log and skip the tiles. If SourceX's program terms change the attribution rule or the payment trigger, start a new cohort series from that date, because earlier and later cohorts are no longer comparable.

Compounding does not relax the fit facts

Compounding never relaxes the basics. Each introduced company needs 50+ full-time employees at peak (contractors excluded), documented operating history and rights to license its records. 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.

The CEO authorizes a licensing discussion and signs for the company, so the CEO must know about and want the conversation. Members ask only metadata questions (what, how much, how long, which system, who owns it) and never ask for or handle records, samples or logins. Whoever submits must have signed the SourceX Channel Partner Agreement first; agreements signed under its earlier title, Channel Partner Non-Circumvention and Non-Disclosure Agreement, remain in force. What a partner is and does is set out in what is a referral partner and what is an AI data partner. For how individual specialists in a network spot candidates, see the industry experts who can find AI data opportunities first.

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 dashboard shows a cohort whose members screen well and whose CEOs want the conversation, register for a partner link, check each company against who qualifies and read the program terms before the next introduction goes in. More guides are in the guides hub.

  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

How many introducers does a network need before cohort signals mean anything?

With only a handful of introducers, a single company moves every rate, so read the case log instead of the percentages. Cohort comparisons start to carry information once each cohort holds several introducers and a dozen or more introductions read at the same age.

Should a network count rewards paid as its main measure of success?

No. Rewards arrive last: only after a completed handoff, the Acquisition Partner's final written acceptance and the other payment conditions. They describe introductions made many months earlier, so they are a lagging check, not a signal of compounding.

What happens when two members of the same network know one company?

Credit goes to the partner whose valid introduction came first, and only if the company submits a verified application inside the 90-day attribution window; later introductions are reviewed as duplicates. Agree inside the network who approaches a company before anyone contacts its CEO.

Sources

  1. refer.sourcex.si/terms
  2. refer.sourcex.si/who-qualifies
  3. refer.sourcex.si
  4. partnerstack.com/resources/research-lab/charts/b2b-saas-compa…

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