App Review Evidence Research

Use public reviews as research leads, not proof of demand, willingness to pay, market size, or a roadmap.

Workflow

1. Define one competitor and one customer job. Do not begin with a feature list.
2. Collect 20 to 50 recent low-rating reviews from the App Store or Google Play. Record links or dates and redact personal data.
3. Copy ../research-template.md and preserve exact excerpts in the raw-evidence table.
4. Cluster reviews only when the user, job, failure, and consequence are materially the same. Keep the count and sample size beside each cluster.
5. For every cluster, state what evidence would disprove it. Exclude vague ratings, one-off bugs, and complaints without a job or consequence.
6. Repeat the check with two to four close alternatives. Distinguish a category-wide failure from one product's implementation issue.
7. Write one narrow hypothesis and interview people who experience the job. Include a smallest test, success signal, and stop signal.

Evidence Rules