A retrospective that asks "what have we earned?" will usually hear zero and draw the wrong lesson. A reward becomes payable only after the Acquisition Partner's final written acceptance of a complete, validated dataset and completed handoff, so most introductions made in the last quarter cannot have reached that point yet. The questions worth asking at the midpoint are operational: did the firm do what it planned, did clients find the topic worth raising, and is the paper you signed still the paper in force?
Month six on the program's three clocks
Three clocks run under every partner firm, and month six lands at a different point on each.
| Clock | What the terms say | Where month six falls |
|---|---|---|
| The agreement | One year; can be terminated on 30 days' notice; renews only with fresh acceptance and signatures | Halfway, with enough time left to change course before renewal |
| Attribution | The first valid introduction within the 90-day window is credited once the company submits a verified application | Introductions from months one to three have closed their windows; later ones are still open |
| Payment | Deadline of 60 calendar days after the later of final written acceptance and completed handoff | Often not reached yet; SourceX publishes no timing averages |
| After an exit | Contacts introduced during the term stay protected for six months after the agreement ends; confidentiality lasts three years | Relevant only if the firm decides to leave |
The format worth borrowing comes from the Army's training doctrine. FM 7-0 describes an after action review as a guided analysis of performance, held during and at the end of an event or operation, whose purpose is to improve what comes next, and it asks reviewers to judge against the standard that was set, using open-ended questions. Month six is the review held during the operation. Month twelve, when renewal needs new signatures, is the one held at the end. Run the first so that the second has evidence to work with. How attribution and contact protection differ is laid out in attribution versus contact protection.
What clients needed, and what the firm said it would do
Start where the program started: the client problem. An established client may hold years of records it created itself that could be licensed for AI training and evaluation, and what it needs is an accurate account of that option from an adviser it trusts, not a pitch. The firm's promised part was small. It would recognize a likely fit, ask the CEO, introduce the company if the CEO wanted that, describe the program accurately and step back while SourceX ran its own review.
Six months in, check whether the part stayed small. Role creep leaves traces in the log: chasing SourceX on a client's behalf about timing, a colleague asking a controller for "a quick sample", a client who heard "you qualify" from someone at the firm before any review had happened. Each is a finding for part E of the template, not a footnote.
The six-month operating review template
Fill in one line per question. Where nobody wrote expectations at month zero, reconstruct them honestly and mark them reconstructed, because hindsight otherwise rewrites them to fit the record.
| Part | Question | Evidence to pull | Verdict to write |
|---|---|---|---|
| A. Record | How many companies did we introduce in each month, and where does each stand at day 180? | The dated name register and SourceX's status for each company | Counts by entry month, open windows shown apart |
| A. Record | How many came back as already known to SourceX? | Duplicate outcomes | A count, treated as a cost of how we searched |
| B. Effort | How many hours went to screening, CEO asks, follow-ups and internal coordination? | Time entries by task | Actual against plan, task by task |
| C. Clients | Did each CEO we approached find it worth raising, and did anything feel like pressure? | Two questions asked after every conversation | Yes and no counts for each question |
| D. Terms | Which terms version and agreement did we sign, and what has changed since? | Our signed copy, any Schedule A, the current terms and their change log | Each change, with its effect on us |
| D. Terms | When does our agreement end, and when must notice be given? | The signed agreement | The end date and the last date for 30 days' notice |
| E. Incidents | Did anyone request records, approach a CEO without the relationship owner, or overstate the program? | Notes and email | Each incident, its fix and whether it recurred |
| F. Expectations | Which month-zero expectations held, missed or went untested? | The month-zero plan or its reconstruction | Held, missed or untested, one line each |
Part D catches what a busy firm misses. The current terms are version 2026-10-22.2, since replaced by version 2026-11-04, and their section 11 lists what changed. From version 2026-10-12, the the greater of $50,000 plus 5% and 15% of Actual Supplier Payout calculation applies to all referrals, including earlier ones. From 2026-10-19, the agreement partners sign is the Channel Partner Agreement. From 2026-10-21, section 3 spells out what counts as an existing relationship, and the list is broad: any lead, account or email correspondence SourceX had with the company before the introduction. From 2026-10-22, section 5 gives the payment deadline and conditions in the words of the agreement's Schedule A, and from 2026-10-22.2 section 4 defines Actual Supplier Payout in Schedule A's words. Signed agreements are never silently replaced, and a signed Schedule A prevails on rate and base. A firm that signed under the earlier title, the Channel Partner Non-Circumvention and Non-Disclosure Agreement, keeps that agreement as signed and is offered a signed addendum that adopts the current text; whether to sign it is a decision for this review. The terms explainer walks through checking which version governs you.
Part A also needs one reconciliation. In PartnerStack and Wynter's survey of 100 senior leaders at B2B SaaS companies with $50 million or more in revenue, 42% tracked partner revenue with multi-touch attribution, 31% with first-touch and 19% with last-touch. SourceX's rule sits at the single-credit end of that range: one partner per company, the first valid introduction. A firm that credits several colleagues internally for one company is keeping a different ledger from the program's, and the two should be reconciled in this review, before anyone argues about who earned what. The campaign reconciliation worksheet gives each company its own 90-day clock.
Active screening or on-request only for months seven to twelve
The realistic choice for the second half is between two modes, with exit as the third option.
| Dimension | Active screening | On-request only | Exit |
|---|---|---|---|
| What starts work | The firm screens its book and asks CEOs who pass | A client raises AI data, or a CEO asks | Nothing new |
| Hours | A weekly budget | Minutes when asked | Closing out open cases |
| Terms | The signed agreement, attribution and duplicate rules all apply | The same | 30 days' notice; contacts introduced during the term protected six months after the end; confidentiality for three years |
| Trust | Highest exposure: the firm raises the topic | Lowest: the client raised it | Explain the change to clients with open introductions |
| Money | Possible, unforecastable; no cap, and the USD 50,000 floor applies once per company | Possible, and rarer | Rewards already earned stay payable when their conditions are met, unless they arose from fraud or a breach |
One gap deserves a written question. The terms say earned rewards survive the end of participation, but they do not spell out how an introduction still in SourceX's review at exit is treated. Ask SourceX in writing before giving notice, and keep the answer with the signed agreement.
Six months at one firm, written up (Illustrative)
Illustrative: an invented 25-person operations consultancy that chose active screening at month zero, budgeted 60 hours for six months and expected "about one introduction a month and a first reward by month six".
| Entry month | Companies introduced | Window closed by day 180 | Status at day 180 |
|---|---|---|---|
| 1 | 2 | Yes | One verified application, in SourceX review; one CEO declined after the call |
| 2 | 1 | Yes | Returned as already known to SourceX through an earlier email; not credited |
| 3 | 2 | Yes | One verified application, in review; one company never applied |
| 4 | 1 | No | CEO call held; application not yet submitted |
| 5 | 0 | Not applicable | No introduction |
| 6 | 1 | No | Introduced last week |
Against month zero, the introduction count held: seven in six months. The "first reward by month six" expectation missed, and could never have held, because neither application has reached a SourceX decision and payment follows acceptance and handoff. Hours ran to 74 against 60, almost all of the overrun in follow-ups. Nine CEOs were asked and seven wanted the introduction; afterwards, eight of the nine said it was worth raising, and one said the follow-up emails felt like pressure. There was one incident: a consultant asked a client's controller for a sample export, and it was stopped before anything was sent. The firm signed under the agreement's earlier title, so part D adds a decision about the addendum.
The verdict: move to on-request only for months seven to twelve, keep the two applications and two open windows moving, and write the renewal test now. The firm will sign again at month twelve if at least one application has reached a SourceX decision and no incident has recurred; otherwise it will let the agreement end without new signatures.
Which client companies belong in the second half
The screen does not change with the mode. A company fits the program's baseline when it reached 50 or more full-time employees at peak, contractors excluded; keeps several years of operating records it created itself; holds the rights to license them; and has a CEO who wants the conversation. The CEO authorizes a licensing discussion and signs for the company, even when a COO or finance lead owns the records and sponsors the idea inside. 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 staffing firm counts only its own full-time employees, not workers placed with clients; describe both numbers and introduce it anyway when its records are strong. A parent and its subsidiaries count as one referred company, so the USD 50,000 floor applies once across the group.
Discovery stays at metadata level throughout: the record type, a rough count, the span of years, the system and who owns it. Files, samples and logins stay out of it; the sample-export request in the worked example is the kind of slip part E exists to catch. What remains unknown until SourceX reviews a company: whether its records fit a current program, whether customer contracts limit reuse and what a license would cover. A company that misses one program may fit another later, so a declined company from the first half belongs on a dated revisit list, not in the bin. For the program's basics in one place, see the partner program overview; for consultants specifically, management consultants.
Pilot, maintain, expand or decline at the midpoint
| Decision | When the record shows | What would reverse it by month twelve |
|---|---|---|
| Expand | Strong CEO answers, hours at or under plan, no incidents and open cases moving | An incident, or hours running past plan |
| Maintain active screening | Clients find it worth raising and hours sit near plan, but outcomes are still open | Open cases sitting with no SourceX decision |
| Pilot a narrower version | Clients answered well, but hours overran or duplicates were frequent | A narrower screen that still overruns |
| Move to on-request only | Hours overran with few CEO yeses, or the firm's attention is needed elsewhere | Clients raising the topic more often than expected |
| Decline and exit | Clients say it was not worth raising, or incidents recur | Nothing needed; earned rewards stay payable under section 10 |
For a firm that never wrote a month-zero budget, the pilot budget with no revenue assumptions is a sound template to adopt for the second half.
Where a midpoint retrospective misleads
- Open is not lost. Introductions inside their window, or in review, are unfinished. Reporting them as failures at month six punishes the most recent work.
- Hindsight edits the plan. Reconstructed expectations drift toward what happened. Mark them reconstructed and weigh them less.
- Seven companies are a small sample. One application reaching a decision can flip every ratio. Report counts, not rates.
- Clients are polite. "Worth raising" overstates enthusiasm, which is why the pressure question is asked separately.
- The program moved under you. Five terms versions dated within October 2026 alone appear in section 11, so a first-half lesson about attribution or payment may describe rules that have since changed.
- Reviewing only what went well. A retrospective that skips declined CEOs and incidents grades the firm on its best cases.
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.
Once the second-half mode is chosen and the renewal test is written down, keep your partner registration current, recheck the companies on your revisit list against who qualifies and reread section 11 of the program terms before month twelve. Other partner decisions are gathered in the guides hub.