What five members can and cannot tell you
The National Center for Complementary and Integrative Health describes a pilot as a small-scale test of the methods and procedures meant for a larger study, run to judge feasibility and acceptability, and it lists estimating an effect size or early efficacy among the ways pilots are misused (our paraphrase of its guidance page). The same discipline fits an association. Your larger effort is offering the topic to every member. The methods are the invitation, the session and the two routes out.
So five participants can answer practical questions. Did the invitation make clear that each member's CEO decides? Did anyone feel pushed? Which questions did the materials fail to answer? When a member said no, did the word reach every staff member and volunteer? Was the hour worth members' time?
Five cannot answer the outcome questions. The share of members that would fit is decided only in SourceX's review, against the criteria of a specific program, and SourceX publishes no acceptance rates or typical payouts. One request for an introduction among five participants is a single event, not a 20% rate, and it should never be reported as one.
The member need and the line the pilot never crosses
Begin with dated member questions. If nobody has asked what licensing company records for AI involves, a short neutral survey comes first; the member survey questions avoid steering answers. Without a stated need, there is nothing for a pilot to test.
Invite every member once, and let members pick themselves. The invitation states the published fit facts so a member can judge whether an hour is worth it: 50 or more full-time employees at the company's peak, contractors excluded; several years of records the company created itself; the rights to license them, including that they have never been licensed for AI training; and a CEO who decides. 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 first five to reply join. Later replies go on a list for a second round. If fewer than five reply, run with fewer and say so in the review.
The boundary does not move during the pilot:
- Nothing about records. The association collects no records, samples, logins, lists of systems, record counts or headcounts, and no result from anyone's self-check. The partner agreement's section 4.2 gives a partner no authority to share a company's records anyway.
- Private self-checks. Members use who qualifies or the company fit checker on their own. A company that applies lists its own systems and years of history to SourceX, not to the association.
- One company, one CEO, one request. An introduction happens only when a member's CEO asks in writing, and the message carries only that CEO's words.
- No partner agreement, no introductions. Section 1 of the terms requires a partner to sign before submitting any introduction. An association that has not signed runs the pilot as education only, and interested members apply directly.
- Money in the open. If the association could be paid, the invitation says so in the body of the message; where the disclosure sits in a newsletter is worked through separately.
The five-member pilot charter
Ten clauses, one owner each. Print it on one page and have the board approve it before the invitation goes out.
| Clause | What it says | Owner |
|---|---|---|
| 1. Purpose | Learn whether a short, neutral session on licensing company records for AI helps members, judged by their own answers | Board chair |
| 2. Who may join | Up to five member companies, each represented by its CEO or someone the CEO sends; the first five replies; anyone may leave without giving a reason | Executive director |
| 3. What the association provides | One invitation, one 45-minute session built from SourceX's published pages, a list of those pages and a feedback form | Program volunteer |
| 4. What it never collects | Records, samples, logins, system lists, record counts, headcounts or self-check results | Everyone involved |
| 5. What can reach SourceX | Nothing, unless a member's CEO asks in writing for an introduction; then one message quoting that CEO | Executive director, and only after the partner agreement is signed |
| 6. Money | Whether the association could be paid if a member it introduced licenses its records, stated inside the invitation | Treasurer |
| 7. Four questions | Was it useful? Did anything read as a recommendation or as pressure? What is still unclear? Should every member be offered it? | Program volunteer |
| 8. Decline and feedback routes | A one-word "no" ends every mention for that member across the association; concerns go to a director who is not running the pilot | Named director |
| 9. Stop rules | Stop at once if a member says its name reached SourceX without its CEO asking, if anyone is asked for records, or if two participants say it read as an endorsement | Board chair |
| 10. Decision | The board reviews in week 6 and picks one: stop, repeat with five different members, or offer to all with the fixes the pilot found | Board |
Clause 7 deliberately asks nothing about the company. Every question is about the session and the invitation, so the answers can be shared with the board without exposing any member. Clause 5 also means nobody tells a participant that a reward or a deal is likely: the agreement's section 4.6 bars a partner from presenting an estimate, positive interest, a signature or an upload as a sure path to a purchase or a reward.
Method note. The charter applies the program terms (Version 2026-10-22.2, since replaced by version 2026-11-04, sections 1 to 3), the partner agreement sample (version 2026-10-19, sections 4.2 and 4.6), SourceX's screening page and NCCIH's guidance on pilot studies, all checked 2026-10-27. It contains no estimate of how many members would fit or what the association might earn, because SourceX has published neither.
A six-week run sheet
| Week | What happens | What is written down |
|---|---|---|
| 0 | The board approves the charter | The signed charter and its date |
| 1 | One invitation goes to all members | A copy of the invitation and the send date |
| 2 | The first five replies are confirmed; later replies are offered a second round | Participant names only |
| 3 | The session | Questions asked, in members' words, with no company details |
| 4 | Private self-checks; a CEO who wants an introduction writes in, or the company applies directly | CEO requests, if any |
| 5 | The four questions go out, with one reminder at most | The answers |
| 6 | Board review | The decision and the reasons for it |
Five members, five endings (Illustrative)
Illustrative. Invented members; each ending is recorded only as the member chose to report it.
| Member | Ending | What the association recorded | What it tells the pilot |
|---|---|---|---|
| A regional distributor | Its CEO asked in writing for an introduction | The request, sent on in one message quoting the CEO | The introduction route worked once; it says nothing about rates |
| An engineering consultancy | Applied directly and mentioned it in feedback | "Applied ourselves" | The direct route is visible, and the association introduced no one |
| A specialty printer | "Not for us" after a private self-check | Those three words; nobody asked why | The boundary held |
| A family-owned food wholesaler | Still deciding at week 6 | Nothing further | Silence is not a yes, and the charter allows no chasing |
| A logistics firm | Said the invitation read like the association recommending SourceX | The comment, word for word | The most useful finding: fix the wording before any wider offer |
The logistics firm's comment outweighs the distributor's request. A wider offer that repeats the wording would put the same impression in front of every member, and clause 9 would have stopped the pilot had a second participant said the same.
Reading benefit, fit and negative feedback
Read the results in a fixed order. Negative feedback comes first: pressure, any sense of endorsement, privacy worries. Route failures come next: did a "no" reach the volunteer who planned to follow up? Usefulness follows, then requests. Leave out any line about members who "qualified"; the association does not know, and only SourceX's review decides. If the board repeats the pilot, compare versions of the wording and the session, not conversion. For the same discipline applied to a commercial partner program, see the pilot with no revenue assumptions.
Where a five-member pilot misleads
- Volunteers are keen. The first five to reply are the most interested members. Their comfort says little about the quiet majority.
- Everyone knows everyone. In a small association a "no" carries social cost. Keep the feedback route anonymous and away from the pilot's organizers.
- Directors' own companies. If a director's company joins, that director stays out of the review and the decision.
- Versions move. Write the terms version on the charter and re-read the terms before any wider offer.
- Five cannot settle a split. If participants disagree, run another five rather than picking the answer you prefer. Whether an unpaid model would sit better with your mission is weighed in the trade association guidance.
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 the board approves the charter and wants clause 5 available, register the association before the invitation goes out; otherwise run it as education only, point members to who qualifies, and keep the program terms version on the charter.