Built it, then rebuilt it the next day
The question
How should a wellness centre reward the gyms and clubs that introduce patients to it?
What I assumed going in
The obvious marketplace shape, which I built without really examining: discount the first consultation to pull the customer in, and pay the partner when a booking happens.
What I actually did
Nothing clever. I described it out loud, and both halves turned out to be pointed at the wrong thing.
A discounted assessment is a less serious assessment. The whole model depends on the first consultation being taken seriously by the patient, and the first thing the scheme did was signal that it was worth less.
Paying per booking pays for bookings, and bookings are fakeable. Paying for something fakeable is a request to have it faked.
What I learned
Rebuilt the next day. The patient pays in full and earns credit toward a programme instead of a discount on the assessment. The partner earns on attendance only, not on sign-up. Only genuinely new patients attribute.
The part I would defend hardest is the partner dashboard, which shows counts and money owed and no patient details at all. That is what makes 'community partnership, not referral commission' an honest description rather than a euphemism — a partner who never learns anything about a patient cannot influence clinical judgement.
Related, same week: a scheme giving money off for a five-star review was designed before anyone checked, and turned out to breach the review platform's policy outright. It became a discount for attending, plus an unincentivised ask afterwards. Buying stars would have been the first dishonest thing on the whole site.
What is still uncertain
Whether attendance-based attribution survives contact with a partner who wants to game it. It has not been stress-tested by anyone trying.
What I would test next
Watch the first twenty partner-attributed patients and look for the pattern a determined gamer would leave: clustered sign-ups, attendance without a follow-up, one partner far ahead of the rest.
It was wrong in a way I could not see until I said it out loud, which is an argument for saying more things out loud earlier.