cleanorlabs/cleanor-search-index

Cleanor Search Index

An open, reproducible monthly index of software-category "popularity" measured from real Google search demand (Keyword Planner historical metrics), by country.

Google Keyword Planner only returns a trailing ~12-month window that rolls forward every month, so a long, comparable history of absolute search demand does not exist unless it is captured prospectively. This dataset does that: it records an immutable snapshot each month and accumulates them into a time series.

• Live charts: https://cleanor.app/trends
• Code + raw snapshots + full methodology: https://github.com/cleanor-app/search-index
• Citable archive (Zenodo): https://doi.org/10.5281/zenodo.21225547

Files

File	One row per	Key columns
popularity-ai-summary.csv	brand × scope (per country + Worldwide)	rank, avg_monthly_searches, share_pct, mom_pct, yoy_pct
popularity-ai-monthly.csv	brand × country × month	year_month, searches

scope is a country code, or WW for the sum across the tracked countries.

Method (short)

For each brand we count only its single head query plus spelling variants of the same name, never nested sub-queries (the head-term volume already contains them, so summing would double-count). Volumes are pulled independently per country, so within-country comparisons are valid. Keyword Planner rounds volumes into bands, so treat values as directional, not exact. Full detail: https://github.com/cleanor-app/search-index/blob/main/docs/methodology.md

Reproducibility

Every derived file here is regenerated from the immutable raw snapshots in the GitHub repo by a credential-free npm run build, and CI asserts the published figures match their raw source.

Citation

Cleanor Labs. Cleanor Search Index. Zenodo. https://doi.org/10.5281/zenodo.21225547

Licensed CC BY 4.0: reuse freely with attribution to "Cleanor Labs, Cleanor Search Index (https://cleanor.app/trends)".