AnswerFix research

How the studies are run

• Businesses are found by asking a search-grounded engine for real local businesses with their own website, then keeping only those whose home page answers.
• Each business gets the same local-intent questions, put to each engine with live web search on, one run per question.
• A business counts as named only on a conservative whole-name match, so a category or partial match never inflates the result.
• Each website is audited against the same readiness checks the free score uses.
• Runs are bounded by a hard engine-call ceiling and checkpointed after every business, so the published sample is exactly what completed.
• The error bars are stated: the measured run-to-run variance appears in every study.

Licence and reuse

Every dataset is published under the Creative Commons Attribution 4.0 International licence. You can quote the findings, republish the charts, reuse the file and build commercial work on it, as long as you credit AnswerFix and link back to the study page. No permission email needed.

Datasets carry aggregates and anonymised per-business rows. No individual business is ever named, in the study or in the file, in either direction.

For live aggregates from checks people run on this site, see the AI visibility benchmarks by industry and city, and for the practical fixes, the AnswerFix guides.