Who comes first in AI answers.
Regular rankings of which brands AI assistants name first, in six countries, with a public method and open data under a CC BY 4.0 licence.
ReadWho makes the press
In the press and in the answers. Do they match?
Every Monday and every month we count which brands are most present in articles, and set them next to what AI assistants answer. Two measures, one question: do they match?
What it measures
Original articles and pickups, kept apart.
A press release picked up by fifty sites is one story, not fifty. For each brand we count original articles and pickups in separate columns, then compare presence in the press with presence in the answers.
The same sample press release picked up by many websites, counted as if every pick-up were a story and counted separating originals from pick-ups.
| Every pick-up a story | Originals and pick-ups separated | |
|---|---|---|
| The count | Fifty articles for the same press release. | One original article and forty-nine pick-ups. |
| Presence | The brand looks like the most present in the press. | The brand has a single story, widely picked up. |
| The comparison | The press seems to say more than the AI answers. | Press and answers are compared on real stories. |
Where the data comes from
From licensed news sources.
Articles come from Webz.io News Search, under licence, and from GDELT, an open dataset. The count uses no AI models: the comparison with the assistants reuses the answers collected in the other series.
Where the articles in the press series come from, what type of source each one is and how it enters the count, without AI models.
| Source | Type | Role in the count |
|---|---|---|
| Webz.io News Search | Licensed API. | Articles with outlet, date and permitted text. |
| GDELT | Open dataset. | International news coverage. |
| Answers from the other series | Observatory collections. | The comparison with what the assistants say. |
Status of the editions
First weekly review in preparation.
The first week of articles is being collected. The comparison between press and answers is published when both measures pass the declared thresholds; until then, no counts and no brands.
Why the press series comes out only when both measures, articles and answers, exceed the declared thresholds, and what happens before then.
Collect the articles
The first week of articles is being collected from licensed sources.
Collect the answers
The answers come from the Observatory’s other series.
Check the thresholds
Both measures must exceed the declared minimum.
Publish the comparison
Only then do brands and counts appear; before that, no numbers.
Related paths
After “Who makes the press”, the next step.
The Observatory publishes rankings with their methodology. From here you move on to the other series, to the edition methodology and to the page that brings them all together.
Every ranking, step by step.
How every Observatory ranking is built: published questions, repetitions, 95% intervals, thresholds, topic sources and corrections.
ReadA topic is rising. Who does AI name first?
Every week, the topics rising in searches and the brands AI assistants name first, country by country.
ReadThe usual questions. Today’s recommendations.
Every month, the questions customers ask before choosing and the brands AI assistants recommend, sector by sector.
ReadBefore buying. Who the assistant recommends.
Every month, the brands and stores AI assistants suggest for the purchase categories with the most interest.
ReadLeading brands. How much room in the answers.
Every month, how much room AI assistants give the leading brands of each sector and country, and with which sources.
Read