[Company]: the one thing I would smoke test first
[One sentence naming the surface and the behavior in question, written so a growth lead can tell in four seconds whether this is worth their time.]
| What I looked at | [Exact surface, for example the mobile product detail page for a specific category] |
|---|---|
| When | [Date, plus device and connection, since these pages change weekly] |
| Funnel stage | [Acquisition, activation, consideration, purchase, or retention] |
| What I do not have | Their analytics, their test history, their roadmap, and any constraint I cannot see from outside. |
What I saw
[Describe the experience neutrally, as a visitor, not as a verdict. What appears, in what order, and what a first-time visitor can tell without tapping anything.]
Why it might matter
[The behavioral reasoning. What a visitor is likely trying to do here, and what this page makes easy or hard. Name any public evidence you have, and say plainly where you are reasoning from pattern rather than data.]
Hypothesis
If [the specific change], then [the specific metric] will improve, because [the belief about how people behave at this step].
The test I would run
Control
[What stays the same, described exactly.]
Variant
[The one change, isolated along a single vector so the result is readable.]
[If traffic to this surface is thin, say so and propose a smoke test instead: a cheaper signal that tells you whether the question is worth a full experiment.]
What this tells you either way
If it wins: [what you would believe next, and what you would test after it.]
If it loses: [what belief about behavior is wrong, and which direction that rules out.]
If it is flat: [whether the change was too small, the wrong step, or the wrong audience.]
Three more I would queue behind it
| Test | Stage | Effort | Score |
|---|---|---|---|
| [Second test] | [Stage] | [Low, medium, high] | [ICE] |
| [Third test] | [Stage] | [Low, medium, high] | [ICE] |
| [Fourth test] | [Stage] | [Low, medium, high] | [ICE] |
This is an outside-in read with no access to [Company]'s analytics, test history, or constraints. The team may well have tested this already and found something I cannot see from here. If any of it is useful, I would rather hear why I am wrong than be right from the outside.