Here’s a workflow I use to go from “maybe” → “evidence”.
Look for:
Workarounds (manual processes, spreadsheets, duct-tape automations)
Non-consumption (people avoid solutions because they’re too heavy/expensive/complex)
High-friction moments (pain spikes in a specific context)
Christensen’s JTBD research suggests workarounds and non-consumption are especially fertile—people are already signaling dissatisfaction.
I like this template:
When [situation], I want to [progress], so I can [outcome].
This keeps you honest about the who and when, not just the feature list.
Avoid: “Would you use/pay for this?”
Use “Mom Test” style questions:
Ask about the last time it happened
Ask what they tried already
Ask what it cost them (time/money/reputation/stress)
The goal is truth, not encouragement.
Pick one:
Fake Door / Smoke Test: landing page or in-product CTA → measure clicks/signups (do it ethically + transparently).
Pretotyping: “fake it before you make it” to test demand/usage without building the full product.
Design Sprint-style prototype + user tests: compress learning into days.
This is basically the Lean Startup idea: maximize validated learning and avoid wasted build cycles.
Ask: “How would you feel if you could no longer use this?”
If 40%+ answer “Very disappointed”, it’s a strong sign you’re on the right track.

Question for IH:
What’s the best validation experiment you’ve run (or wish you ran earlier)?