The origin
RestoRank is led by Nicolas Spielmann, who operates IMA, a restaurant-club in Val d’Isère. Guests started telling us they’d asked an AI where to eat, then booked through an aggregator instead of our own site. So one question became obvious: when someone asks ChatGPT where to eat, who comes up, and where does the booking go?
No tool answered that question for restaurants. So we built one, from the floor, not from an SEO agency.
Our take
RestoRank isn’t an SEO tool with a restaurant skin on top. We built it for the trade, not the other way around:
- what’s on the plate and what happens in the room, not just keywords;
- hyperlocal discovery and the walk-in guest;
- the commission war: direct booking vs aggregator;
- the words diners actually use, which AIs pick up too.
How we measure
We query the five big AIs (ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews) through their official APIs, with the actual questions diners ask. We don’t give opinions, we measure: mention rate, average rank, share of voice against competitors, and which channel ends up with the booking. Data, tracked week after week.
We use it on ourselves
We run RestoRank on IMA continuously, before we offer it to anyone else. That’s what keeps the tool useful to a restaurateur in service, instead of pretty on a dashboard.