inApp

The user voice, not a summary

Reviews

All 1,451,072 reviews are here — and every one is labelled. A specific topic is assigned only when the text contains an explicit signal; short or ambiguous reviews stay in an honest sentiment category.

71
niches
4,623
apps
225
translated niche patterns
1,451,072
labelled reviews

Data passport

What is available and labelled

Updated August 13, 2026

Source texts

All 4,623 apps can be opened by category

1,451,072 / 1,451,072

A label on every review

Exactly one topic or honest sentiment category per text

1,451,072 / 1,451,072

Specific semantic topic

37.9% of the corpus · explicit signals only

550,502 / 1,451,072

Themes of individual apps

29 niches · 19.9% of the complete corpus processed

937 / 4,623

Live evidence

Every pattern includes quotes, app name, and rating.

Three analysis layers

Review labels, niche patterns, and deep product topics stay separate.

Honest remainder

Short or ambiguous texts are not turned into invented insights.

How the labelling works →

Explore the corpus

Find a niche

71 results
Sort

The same data is available to your AI agent: the inApp MCP server