Signal Census publishes category reports on scraping and automation tools. Each report applies the same public rubric so reports can be compared across categories and quarters.
This is an editorial scoring system, not an academic benchmark. It is designed to be transparent, reproducible in broad outline, and open to correction.
What is in scope
A report covers one product category for one quarter. A product is in scope if it is a usable commercial or sustained-free product (hosted service, marketplace listing, or API) within the category, generally available at the time of census. Bare open-source repositories that require the user to self-host are excluded.
The specific inclusion rule for each report — which category, which time window, what demand threshold — is stated in the report’s “How we scored this” box.
Where candidates come from
For each report we search a fixed set of public channels: the Apify Store filtered by category keywords, vendor directories (G2, Capterra, AlternativeTo), vendor websites found via category search terms, recent Reddit / Hacker News / forum threads, and the previous report in the category. The full list of candidates — kept and excluded, with reasons — is published as part of the report’s dataset.
How tools are scored
Each in-scope product is scored on a 0–10 scale across six axes:
- Coverage — does it address the category’s full range of sources, or only a slice?
- Extraction quality — does it return complete and accurate records against a sample of test queries?
- Reliability — does it complete runs without silent failures, partial output, or undocumented limits?
- Freshness and maintenance — is the product actively maintained and responsive to source-site changes?
- Documentation and usability — can a technical buyer evaluate, configure, and operate it from public docs alone?
- Pricing transparency — is the cost predictable in advance?
The overall score is the unweighted arithmetic mean of the six axis scores, rounded to one decimal. Equal weighting is deliberate: the right weights depend on a given buyer’s priorities, and a fixed weighting would hide that subjectivity behind false precision. The per-axis scores are published so readers can recompute with their own weights.
Conflicts of interest
The publisher of Signal Census operates commercial products in several of the categories covered, and may have affiliate relationships with vendors in others. Where a report covers products in which the publisher has a commercial interest, the report carries a disclosure box naming the relationship and the affected products.
Conflicted products are not excluded from rankings. They are scored with the same rubric as everything else, and the full scoring data is published so readers can verify — or challenge — the result.
Corrections
Readers may submit corrections via /contact. Accepted corrections — including scope refinements, score changes, or new evidence — appear in the per-page changelog at the bottom of each affected page.