The Hidden Commission Economy: What AIs Really Cost Your Margins

Let's talk money. An AI recommendation that feels like a win can, in practice, cost you margin, when it pushes the guest to book through an aggregator.

The double toll

An AI recommends your restaurant. Good. But to book, it cites a link. If that link points to a booking platform, you pay a commission per cover. The guest was already coming. You are paying for a cover you had already earned in the dining room.

In a large share of restaurant answers, the cited link points to an aggregator rather than the restaurant's own site. Put plainly: by default, "AI distribution" runs through paid intermediaries.

Run the numbers

Take your monthly covers from online bookings, the share that goes through an aggregator, and the average commission rate. The annual figure tends to sting. Every percentage point you pull back into direct booking lands in your margin, not someone else's.

That is why RestoRank measures the own site vs. aggregator ratio outright: of AI answers that cite a source, what share sends traffic to you?

Pull the direct link back (without declaring war on the platforms)

The point isn't to bury aggregators. They still pull weight. The point is to be citable next to them, so the AI has a direct route to offer:

1. Full schema.org/Restaurant structured data (menu, FAQ, reviews, opening hours).
2. Menu in HTML, readable by crawlers. Not a PDF.
3. Reviews collected on your own site (official widget), on top of the platforms.
4. Direct booking clearly visible, with a clean link.

A fight hospitality runs differently

Hotels have been fighting OTAs on rates for twenty years. Restaurants have their own math: per-cover commission, tight margins, volume. A tool built for hotels doesn't think in those terms. RestoRank does. For a restaurant, pulling the direct link back means defending your margin one plate at a time.

Find your leak first. The rest follows.