Original Data, Honestly Reported
Most SEO "studies" are marketing with a chart on top. We publish the query set, the method, and the limitations — including the ones that weaken our own conclusions. If a sample is small, we say so rather than rounding it into a headline.
What we've measured so far
Who Google's AI Overviews Actually Cite
We pulled the AI Overview for 40 SEO-related queries and recorded every source cited, then compared them against the organic results for the same query. Nearly half of all citations went to sites that weren't ranking on page one — and the most-cited domain wasn't a publisher at all.
- 48% of citations came from domains outside the top 10
- Commercial and tool queries showed an AI Overview far less often than informational ones
- YouTube was the single most-cited domain, ahead of every SEO publisher
Rules We Hold Ourselves To
- Publish the method — data source, date, location, and sample size, every time
- Publish the limits — including the ones that cut against our conclusions
- No inflated samples — a small honest study beats a large vague claim
- Share the raw set — ask and we'll send the query list so you can replicate or extend it
- Separate finding from opinion — what the data shows, then what we'd do about it
Why we publish this at all
Partly because it's how we actually make decisions for clients — we'd be running these pulls regardless. Partly because an agency that only publishes advice is asking you to take its word for things.
If you'd like us to look at a specific question, or want the underlying data from any study, tell us. We're more interested in being useful than in being quoted.
Want This Applied to Your Site?
The free audit uses the same data sources on your pages — and shows what to fix first.
Get a Free Audit →