Quick answer
In one hour, a PM can go from “random interesting threads” to a short stack-ranked list of problems with quotes, frequency notes, and explicit unknowns. The trick is fixed timeboxes, a single note template, and stopping when you have three decisions: what to validate in sales calls, what to spike in engineering, and what to ignore for now.
Best for: B2B PMs at startups where buyers discuss work in public (dev tools, data, security adjacent, ops) - not for PMs whose ICP never posts about work online.
Before you start (5 minutes)
- Hypothesis: One sentence (“We think onboarding fails on X for team Y”).
- Scope: 2–3 communities your ICP actually uses (do not boil the ocean).
- Output: One page in your doc tool with sections: Top problems, Counterevidence, Next steps.
Minute 0–15: Capture raw signal
- Run narrow searches on the problem domain (product name, workflow, competitor names).
- Save 10–15 links max; prefer threads with specific constraints (budget, stack, timeline).
If you use Needle, run Search across the communities your plan includes and export where your plan allows - Growth and Scale include exports (see Pricing); Solo does not include export capability per current plan config.
Minute 15–35: Triage with a simple scorecard
Score each thread 1–3 on:
| Criterion | Question |
|---|---|
| Urgency | Is someone blocked today? |
| Fit | Is this our ICP or a misfit persona? |
| Actionability | Can we ship or message something concrete in weeks? |
Drop anything that scores 1 on fit. Keep disagreements as counterevidence, not noise to delete.
Minute 35–50: Extract buyer language
Copy verbatim phrases that could land on a homepage or changelog - especially verbs (“rip out”, “reconcile”, “rollback”). Tag each quote with persona and environment (self-serve vs enterprise-y language).
Minute 50–60: Decide interview vs search vs ship
- Interview if stakes are high and posts are thin.
- More search if you see the same symptom words but not root cause.
- Ship/spike only when multiple independent threads show the same workaround.
“Best for” recap
| Situation | Use this hour |
|---|---|
| Post-launch confusion after a release | Yes - fast sentiment triage |
| Brand-new market with no vocabulary | Partial - pair with expert calls |
| Executive asks for “what customers want” | Yes - as a structured briefing |