Reports
AI visibility reports, with the data open
Reports from the AI Recommendation Index show which brands AI engines recommend, by industry and country. We also repeat studies of claims made about AI search. Every figure comes with its sample and range, and every report with its data.
Published
Nothing published yet
The first report follows once the Index has enough data.
A report needs repeated sampling over several weeks before its numbers mean much, so we won't publish one built on less. Until then, the AI Recommendation Index and its methodology are open to read.
Planned
What we'll publish
Monthly Index report
Which software, products and services AI engines recommend most, by category and country, with the change since last month. A change is only called a change when it's bigger than normal variation.
Industry reports
One industry at a time (software first, then retail, travel, finance and health), across every market we sample. Brand and organization level only: we report what engines said, never advice.
Seasonal reports
Shopping-season and year-start editions, published from the same open panel, so they compare like for like with the monthly data.
Replication studies
We re-measure published industry claims (how fast AI citations change, how often AI lists overlap, how consumer apps and APIs differ) and publish the result with intervals, whichever way it comes out.
Every report
What every report includes
- The sample: how many answers, from which engines, in which markets and over which dates.
- A 95% interval on every share, and the prompt list we asked.
- The data as a CC BY 4.0 CSV, free to reuse with a credit line.
- A link to the methodology version it was measured with.
Related