The form was never the reason fund managers stopped sharing.

Preqin sells access to data that fund managers supply for free, so the product has a supply side and a demand side and only one of them pays. When fund managers stopped contributing performance figures, every subscriber got less back for the same money. That is a renewals problem arriving through a channel nobody was watching.

The obvious diagnosis was friction: submission is tedious, so make it easier. Six options were on the table, and the three easiest to engineer all attacked collection. Research established that collection was the problem nobody had. Partners weren’t withholding numbers because the form was hard. They were withholding numbers they were embarrassed by, and no upload tooling fixes embarrassment.

So the work had to move from collection to exchange, and that reframe was not free: it argued the team off the three cheapest options and onto the three that required deciding what a fund manager gets back for a number that might make them look bad. My first answer to that was a leaderboard. It backfired, and it cost partner trust to find out. The version that worked kept the comparison and removed the identity. Contribution retention went from 5% to 15%, and diligence cycles ran four months instead of ten.

Fund managers weren’t contributing performance figures, reducing the platform’s value for every subscriber who relied on it. Revenue and renewals were exposed.

The obvious diagnosis was friction. Submission was tedious, so make it easier. Research pointed somewhere else.

That last question changed the brief: what does a fund manager get for sharing a number that might embarrass them?

Withholding has two possible causes and they want opposite products. If the cost of contributing is effort, better tooling is the answer and the scraper is the best of the six. If the cost is exposure, tooling makes submission faster for people who were never going to submit, and the honest version of shipping it is that the team gets a completed roadmap item and the data still doesn’t arrive.

Every signal in the research pointed at exposure. “If my numbers aren’t great, I don’t want anyone to know” is not a sentence about a form.

The uncomfortable part is that this argument was expensive to win. Collection was the cheaper build, the more legible commitment, and the one that needed nobody to agree on what a contributor gets back. Moving to exchange meant asking for engineering time against a benefit I could only argue for, and I was about to get the first version of that benefit wrong.

The durable output is not the inline field. It is the cohort floor: four funds before a benchmark can be shared, eight before a median. That rule is a control written into the product rather than into a policy document, and it is what lets anonymity survive contact with the thing that makes the data useful, which is slicing it finely enough to be relevant. Somebody has to keep deciding where that floor sits as the dataset grows, so it is a governance commitment as much as a feature — and it is the piece most likely to still be load-bearing in the product now.

The second is a policy rather than a screen. The form ships short and gets extended from what people actually complete, not from what an internal advocate can justify. That converts a quarterly argument into an evidence question, which is the only version of that argument anyone can win.

What it did not unlock is a settled answer on the leaderboard, and the note below is about that.

Both causes produce the same observation — submissions are down — and the roadmap options that follow from them share no overlap at all. One is a tooling problem and one is an incentive problem, and the first is always cheaper to build, easier to scope, and more comfortable to commit to in a planning meeting.

What I do differently since: when a contribution rate falls, I try to establish which of the two it is before any option gets estimated, because the estimates themselves bias the choice. Three of our six options were cheap precisely because they assumed the easier cause, and a cheap option that assumes the wrong cause is not cheap.

The second half of the lesson is in the note, and it is the harder one. Getting the diagnosis right did not stop me getting the first remedy wrong, and the research had already told me why.