A network expansion plan with clear quality gates: five stages from a founding cohort to open enrollment, each opened by evidence and guarded by a rollback trigger

Short answer

Expand a referral network one stage at a time and let evidence, not sign-ups, open each gate. A workable plan has five stages, from a founding cohort to open enrollment. Each stage opens only when the previous cohort's introductions were clean, active companies fit within the hours available and the duplicate register holds no unresolved claims; each also carries a rollback trigger that sends the network back a stage. Logo counts open none of the gates.

Visual briefFull-size diagram
Five stages in a row, founding cohort, same segment, adjacent segment, new region and open enrollment, separated by gates that check quality, capacity and duplicates, with a rollback arrow returning from the new-region stage to the stage before it.
The stage-gated playbook on this page. Gate thresholds are set by the network in advance; SourceX publishes none.

Every new cohort adds introducers before it adds qualified companies, and the briefing, screening and overlap work arrives first. The plan below exists so that growth never outruns the people checking it.

Activation and referral quality, counted per cohort

Count companies, not partners, and give each measure a denominator that keeps quiet members in view. One row per company, with a corporate group kept on one row because the reward cap applies once across a parent and its subsidiaries.

MeasureNumeratorDenominatorWhat it tells you
Activation rateIntroducers with at least one ask that passed the pre-check within 90 days of joiningEveryone who joined the cohortWhether the cohort engages, counting the silent
Clean-submission rateCompanies submitted with CEO consent, every known fit fact passing and no duplicate foundCompanies screenedWhether screening works
Duplicate rateSubmitted companies later found to have earlier SourceX contact or a competing introductionCompanies submittedWhether segments overlap
Correction countStatements to companies that had to be corrected in writingActive introducersWhether briefing holds up in the field
Company feedbackCEOs who said the introduction was clear and unpressuredCompanies introducedThe company's view of the partner

Vendor maturity models often start elsewhere. PartnerStack's four-stage framework (exploring, scaling, building, optimizing) was built from conversations with three of its customer success managers and a study of real programs, and lists exploring-stage KPIs such as clicks and signups, partner onboarding time and first deals sourced. Those suit a software affiliate program. In a program where the company applies itself and SourceX's review decides, clicks and signups describe reach and say little about whether the next cohort will bring clean introductions. Company feedback is the measure most networks skip; a six-question version is in tracking company satisfaction before asking for another introduction.

The stage-gated expansion playbook

Five stages. A stage opens only when every entry condition holds, and it closes to new recruits as soon as its rollback trigger fires. The thresholds are examples for a network to set in writing before the stage starts; SourceX publishes none of them.

StageWho joinsEntry gate (all must hold)Capacity gateRollback trigger
0. Founding cohortFive to eight introducers from one segment, known personally to the program leadEach has signed the SourceX Channel Partner Agreement and passed training on terms version 2026-10-22.2, since replaced by version 2026-11-04Active companies fit the program lead's weekly hoursNone: this stage is the floor
1. Same segment, widerUp to ten more introducers from the same kind of firmStage 0 produced at least three clean submissions, no unresolved duplicates and no uncorrected claimsCeiling recalculated from stage 0's real time logDuplicate rate above one in five for a quarter, or capacity exceeded two weeks running
2. Adjacent segmentA neighboring profession, such as transaction accountants after M&A advisorsA written profile showing the segment's clients commonly pass the size and history checks, and stage 1 activation of at least one in threeA second person trained to screenClean-submission rate below half of stage 1's after two quarters
3. New region or languageIntroducers whose clients sit outside the U.S., Canada and the EUThe segment profile checked against the location and English-language ruleScreening help for mixed-language recordsMore than a third of screened companies fail on location or language
4. Open enrollmentAnyone who registersSelf-serve training with scored skill checks, and written answers to the most common questionsSupport questions answered within five working daysNew-member activation below the floor you set, or support waits above the limit

Notes on the stages that most often go wrong:

  • Stage 2. SourceX's home page lists the professions it considers best placed to refer: M&A advisors and business brokers, private equity and operating partners, fractional CFOs and accountants, IT managed-service providers, wind-down and restructuring advisors, and operators and executives. Moving between neighbors on that list is the natural adjacent step, and the riskiest, because neighbors share clients. If the new segment is private equity operating partners, expect introductions framed around operating improvements, the lens set out in value creation without multiple expansion.
  • Stage 3. Location changes the fit test itself. 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. A segment profile for this stage has to estimate the language of clients' records, not just their headcount; the English-language eligibility guide sets out how that rule is read.
  • Stage 4. Open enrollment removes the program lead's personal filter, so the training and the written answers have to do the filtering instead. Do not open it while any earlier stage is rolled back.

Map capacity, support, concentration and overlap before each gate

Four quantities decide whether a gate should open, and each is cheap to estimate before a cohort starts.

Capacity. The active-company ceiling is the weekly hours available for active companies divided by the average weekly hours one active company takes. With 16 hours and an average of 1.6 hours per company, the ceiling is 10 companies (Illustrative). Recompute the average from a real time log after each stage; the first estimate is always a guess.

Support. New introducers ask most of their questions in their first month. Budget answer time per cohort before it joins, and treat a growing wait as a capacity failure, not a communication problem.

Concentration. Track the share of clean submissions from the top introducer and from the top segment. If one person or one profession supplies most of them, the network's growth depends on a single relationship base, and stage 4 should wait.

Overlap. Segments that serve the same mid-market companies, such as brokers, accountants and lawyers, will reach the same CEOs. Two segments cannot split one company, because the terms credit a single partner, the one whose valid introduction of the new company landed first inside the 90-day attribution window, and that credit is settled when the company submits a verified application. A company SourceX already knew through an earlier application, relationship, lead, account or email correspondence is a duplicate for every segment. Before a new segment joins, run its likely client names against the existing register at metadata level, and keep one register across all segments. The register mechanics are in preventing duplicate claims before they reach SourceX.

Three cohorts at the gate (Illustrative)

Illustrative. The networks, cohorts and counts are fictional; they show how each gate behaves.

A region cohort that underperformed. A network of advisors opens stage 3 with eleven introducers in Singapore and South Africa. In two quarters they screen 24 companies; 10 fail because a large share of each company's records is in languages other than English, so the companies fall short of the 90% rule. That is more than a third, and the rollback trigger fires. The network stops recruiting in stage 3, keeps the eleven introducers, and rewrites the segment profile to ask about record language before headcount.

A duplicate risk between segments. At stage 2, an accountant and a business broker in the same network both know the CEO of a 400-person distributor. The shared register shows the broker logged it three weeks earlier, so the accountant stands down. The earlier-contact question then reveals that the distributor's finance team had exchanged emails with SourceX months before either of them, which makes it a duplicate: neither is credited. Separately, the distributor's parent owns a second operating company that a partner from another network introduced last year. Whether that history bears on credit for anything else in the group is not something the published terms decide; it is a matter for SourceX's duplicate review, and the plan records it as unresolved rather than as a pipeline gain.

A capacity limit at stage 1. The stage 0 log puts the ceiling at 10 active companies. Two months into stage 1, 13 are active and the program lead is working weekends to keep up. The capacity gate fails, recruitment pauses, new candidates wait in a dated queue, and stage 2 is postponed until a second screener is trained.

Correction paths that keep expansion honest

Growth plans usually describe the way up. Write down the way back too.

  • Rolling back a stage. Stop recruiting into it; do not offboard the introducers already in it. Fix the cause and re-run the gate after a quarter.
  • A new terms version. The terms say signed agreements are never silently replaced, so introducers with a signed Schedule A keep its terms. Re-brief every cohort on the new version and record which version each introduction was screened under; a cohort report for referrals made under different criteria shows how to compare like with like.
  • A claim corrected in the field. When an introducer has told a company something the program does not offer, the correction goes to the company in writing, and the briefing that allowed it is fixed for the next cohort.
  • Offboarding. The SourceX Channel Partner Agreement runs for one year, can be ended on 30 days' notice and protects qualifying contacts introduced during its term for six months after it ends. Rewards already earned stay payable once their conditions are met, unless they arose from fraud or a breach. The terms say nothing specific about introductions still in review on the day a partner leaves, so agree how those will be handled with SourceX in writing before offboarding anyone, and tell the departing partner only what has actually been agreed.

Where a stage-gated plan goes wrong

  • Thresholds set after the results. A gate written once the data is in will always open. Date the thresholds before each stage.
  • Small cohorts. One strong introducer can carry a cohort of eight through a gate on luck. Look at the spread across introducers, not only the totals.
  • Moving criteria. SourceX applies the criteria of a specific program, and a company that fails one program may fit another later. A segment profile is valid on the date it was written.
  • The cost of caution. Gates slow the growth of good segments too. That is the price of never having to explain a wave of poor introductions to companies that did not ask for them.

The rest of the partner-network playbooks are collected in the guides library.

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.

Write your stage 0 entry gate today, brief the founding cohort from who qualifies and the program terms, and have each founding introducer register for a partner link before the first ask.

  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 should a founding cohort have?

Few enough that the person running the program can screen every company personally, usually five to eight. The point of the first stage is a time log and a clean register, not reach.

Should a network roll back if one stage goes badly?

Roll back recruitment, not people. Stop adding introducers at that stage, keep supporting the ones already active, fix the cause and re-run the gate after a quarter.

Do the gates change when SourceX changes its criteria?

The gates stay; the briefing changes. Re-brief every active cohort on the new version, record which version each introduction was screened under, and compare cohorts only within the same version.

Sources

  1. refer.sourcex.si/terms
  2. refer.sourcex.si/who-qualifies
  3. refer.sourcex.si
  4. partnerstack.com/articles/four-stages-of-partnership-maturity…

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