From Finding to Hypothesis
// gate two, continued
"The size guide is buried" is a finding. It is not something you can run, win, lose, or learn from.
A hypothesis is a finding plus a commitment. It names what you'll change, what you expect to happen, and why you believe it — and it does all three before you look at the result, which is the only time any of it is worth anything.
the shape
| we believe | the change — specific enough that two people would build the same thing |
| will | the effect, on a metric you already measure |
| because | the evidence from the snapshot, with the number in it |
Worked example, from the footer and exit-path cards:
We believe putting delivery date and free-returns copy next to the add-to-cart button will raise add-to-cart rate because 193 footer clicks and 68 exits to the shipping page say people need that answer before they'll commit, and right now they have to leave to get it.
what disqualifies one
- "test the headline" — no change named, no effect named, no evidence. This is a to-do, not a hypothesis
- "improve the gallery" — two people would build two different things
- "make the page better so it converts more" — the effect restates the goal instead of predicting a mechanism
- "add social proof because social proof works" — the evidence is a blog post, not your snapshot
- a bundle of six changes with one expected effect — you'll learn that something worked
The last one is the expensive mistake. Every change you bundle in buys you a little more chance of a win and costs you all of the explanation. Six changes and a 9% lift tells you nothing you can reuse on the next page.
one more test
Before it goes in the backlog, ask what you'd conclude if it lost. If the honest answer is "I'd assume we built it wrong", it isn't a hypothesis yet — you don't believe it enough to be surprised, so there's nothing to learn either way.
// your three hypotheses
Take your three strongest findings from the last page and write each as one. Keep the "we believe / will / because" shape — it's doing work.
These three follow you through the rest of the session — you'll score them, work out whether you can prove them, and leave with them in order. Short and specific beats thorough.