MIA-FR · J0 wave · Media kit
AI visibility is not a single ranking.
On the same corpus, Edikka citation rates range from 0% in ChatGPT to 40.28% in Google AI Mode: each environment must be measured separately.
- Published dataset
- 2026-08-26
- Version
- 1.0.0
- DOI
- 10.5281/zenodo.22114665
- Licence
- CC BY 4.0
Editorial angle
The useful number is not “my AI rank”, but a dated series of mentions, citations, sources and downstream outcomes.
« A serious measurement keeps the zeroes, technical blocks and platform differences. »
01 · Three figures to retain
AI visibility is not a single ranking.
02 · What the data shows
Citations by platform
03 · Method in four points
Method in four points
- 25 non-branded French prompts, four platforms and three repetitions: 300 planned runs.
- Five technical blocks retained and excluded from denominators: 295 valid observations.
- Mentions, citations, source domains and URLs recorded separately.
- 60 responses blind double-coded; prompt-clustered bootstrap with 10,000 replications.
04 · Limits to quote
Limits to quote
- One brand, one collection date and a French-language corpus.
- Descriptive snapshot, not market share, causality or a forecast.
- Generative answers vary by service, session, location and date.
05 · Two publication-ready visuals
Two publication-ready visuals
NLM protocol record. This external record covers the MIA-FR protocol, not the J0 results. Its inclusion, derived from Zenodo metadata, is neither an evaluation nor an endorsement by NLM or NIH.
Recommended citation
Morel, B. (2026). MIA-FR · J0 wave: Edikka visibility in AI responses in France (26 August 2026) (Version 1.0.0) [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.22114665
The key-figure CSV and visuals are released under CC BY 4.0. Credit “Source: Edikka” and link to the DOI. Do not infer individual results from aggregate data.
Need a method check, a chart or an interview?
Bertrand Morel, Edikka’s founder and editorial director, answers questions directly.