Every research post-mortem I’ve sat through starts with the same question: was the sample right, were the questions right, did we talk to enough people. After enough of these, I started tracking a different number — not how many sessions we ran, but how many of our findings anyone could still name three months later. The number was small enough to embarrass me.

The interviews were never the problem

Six studies, one quarter, decent recruiting, competent moderation. When I went back through the recordings looking for what went wrong, I mostly found good research. Participants who contradicted themselves in useful ways, moderators who followed up on the right things, real signal buried in real conversations. The problem wasn’t upstream of the readout. It was the readout itself, and everything downstream of it.

A readout deck is built to be presented once, to a room, live, with someone narrating the slides. That format has a hidden assumption baked into it: that the value of the research gets transferred in that one meeting, to whoever’s in the room. Everyone who wasn’t there — which by month three is almost everyone — gets nothing but a file they’ll never open, because a deck without its narrator is mostly blank space and quotes with no context.

What actually gets reused

I started checking, deliberately, what research artifacts people actually reopened later — not what they said was useful in the debrief, but what showed up in a Slack search three months on. It was never the deck. It was two things: a single sharp quote someone had pasted directly into a ticket, and a short written summary that read like a memo, not a presentation.

Both of those survive without a narrator. A quote is self-contained — it doesn’t need thirty minutes of context to land. A memo is built to be read cold, by someone who wasn’t in the room, at 11pm before a decision the next morning. A slide deck is built to be watched, once, with someone standing next to it. Almost none of our findings needed to be watched. They needed to be read, later, by someone who didn’t know the study existed.

What changed

The deck didn’t disappear — some decisions still need a room and a live conversation. But it’s no longer the primary artifact. Every study now produces a one-page written memo first, addressed to nobody in particular, built to make sense with zero shared context. The deck, when we still make one, gets built from the memo, not the other way around.

The other change was smaller and mattered more: every finding gets a home outside the research doc. A quote that should shape a specific feature gets pasted into that feature’s ticket, not left in a shared drive folder three clicks from anyone who’d need it. This felt like busywork the first few times. It’s the only part of this that’s actually changed what ships.

Rhys O’Sullivan is a product designer at Fieldstone, previously Loop. This is one in an occasional series of notes on research, systems, and the process of making things.