Validating a startup idea by searching only Reddit - or only Twitter - is like interviewing one person in a crowded room. You get signal, but not a representative picture.
This guide explains what each platform is good for, a repeatable workflow, and when to move from manual tab-hopping to a unified search pass. For product specifics, see Search documentation and plans and limits.
The blind spot problem
| Platform | Typical signal | What you miss if you stop here |
|---|---|---|
| Raw pain, alternatives, rants | Enterprise buyers on SO/HN | |
| Hacker News | Founder evaluation, Show HN feedback | Non-technical users |
| Stack Overflow | Error messages, SDK confusion | Emotional switching stories |
| GitHub | Issues, dependents, maintainer burden | Consumer prosumer threads |
| YouTube | Tutorial pain in comments | Written long-form nuance |
Example: You find Reddit threads complaining about manual podcast editing. YouTube comments reveal 10× more volume on clunky export workflows - a positioning angle you would miss on Reddit alone.
Workflow: one query, four passes
1. Pick 3 phrases (alternatives / pain / constraint)
2. Run each phrase across Reddit + HN + SO + GitHub
3. Log 10 quotes with URL + date
4. Cluster by job-to-be-done
5. Reply on 2 threads before updating landing copy
Time box: 45–60 minutes per phrase cluster. Quality of quotes beats quantity of tabs.
Platform selection by ICP
Developer tools: Stack Overflow + GitHub issues first, HN second, Reddit for self-hosting and "alternative to X."
Prosumer SaaS: Reddit + YouTube comments; HN if technical buyers exist.
Vertical B2B (legal, finance, health): Niche subreddits + industry forums - verify rules before posting (Reddit checklist).
Early idea validation: Start with free Trending Problems for category momentum, then drill into threads with search.
Emotional context beats keyword volume
Keyword tools show what people type into Google. Community threads show how they feel when something breaks - frustration, urgency, budget shock.
Prioritize threads with:
- Explicit timeline ("need this before launch")
- Switching language ("moving off X")
- Comparison requests with constraints ("team of 5, SOC2")
See alternatives and switching intent for query templates.
Competitor and category monitoring
The same workflow finds:
- Where incumbents are mentioned negatively
- Feature gaps users describe as workarounds
- Positioning phrases competitors own ("easier than X")
Run monthly, not only at launch. Pair with conversation demand vs SEO so community language feeds comparison pages when you have GSC footholds.
When to automate
| Stage | Approach |
|---|---|
| First 10 customers | Manual search + spreadsheet |
| Weekly repeat queries | Saved searches or unified Search |
| Brand + competitor | Scheduled monitoring (Auto Search on paid plans) |
Needle Search runs one query across communities on your plan with intent-ranked results - useful when step 2 above becomes a weekly habit.
Common mistakes
- Platform tourism - posting everywhere, engaging nowhere
- Ignoring culture - HN comment tone ≠ Reddit comment tone (HN playbook)
- Treating stars or upvotes as revenue - engagement ≠ pipeline
- Skipping Stack Overflow for devtools - buyers live in error messages