Five Letters on P.E.D.A.L.

Dear Operator,

There is a quiet thing happening to your restaurant right now, and you cannot see it from inside the dining room.

A machine is reading you. Not one machine — many. Google's AI Overview, ChatGPT, Perplexity, the recommendation engine inside DoorDash, the model behind a guest's voice assistant. They are reading every digital surface you have left in the world: your Google Business profile, your Yelp listing, your Instagram bio, your website, your delivery app description, the language inside your reviews.

And here is what I have learned after analyzing more than two hundred restaurants: those surfaces almost never agree.

Your website calls you upscale farm-to-table. Your Yelp says casual neighborhood spot. Your Instagram bio says modern American. Your DoorDash listing says comfort food. Each of these was written by a different person, on a different day, for a different reason. None of them was wrong. Together, they are a contradiction the machine cannot resolve.

The machine does not pick the best version of you. It averages them. Or it picks the weakest one. Or, most often, it lowers its confidence in you and recommends someone else — someone whose signals all say the same thing.

Platform consistency is not about identical copy. It is about coherent identity. Every surface telling the same story about who you are, what you serve, and who you serve it to.

The fix is unglamorous. Pull up your three loudest surfaces side by side. Read the first sentence of each. Ask one question: do they describe the same restaurant?

If the answer is no, you have not been outvoted. You have been misread.

— Margaret

Dear Operator,

I want to tell you about a sentence that does nothing for you.

Great food, great service, great atmosphere.

It is true about your restaurant. It is also true about every restaurant. Which means, to a machine reading the internet at four o'clock on a Friday afternoon trying to answer the question where should I take my wife for our anniversary, it is the same as silence.

Experience clarity is the discipline of writing language so specific that a machine can quote it back. A quiet, candlelit room where the tasting menu unfolds over two unhurried hours. Twelve tables, hand-rolled pasta cut every morning, a wine list that leans toward Piedmont. A corner banquette built for long conversations.

These are sentences a machine can use. They give the model something to match against intent. Anniversary becomes candlelit. Business dinner becomes quiet corner. Birthday becomes long table, loud room, shared plates.

Generic descriptions are not neutral. They are demoting. The machine reads them and concludes you are interchangeable. Interchangeable means invisible.

Try this tonight. Cover the name on your website's homepage. Read the first paragraph aloud. Ask whether someone who had never visited could pick your restaurant out of a lineup of ten in your category.

If they could not, the problem is not your food. The problem is that the machine has nothing distinctive to remember you by.

— Margaret

Dear Operator,

There is a gap between the way you talk about your restaurant and the way your guests search for one.

You write elevated cuisine, artisanal preparations, curated wine program. They type anniversary dinner downtown, quiet restaurant for a business meeting, date night with good cocktails, birthday spot for a group of eight.

Two languages. Same room. Almost no overlap.

Discovery language is the bridge. It is the work of taking the moments your restaurant actually serves — the celebrations, the first dates, the deals closed over a second bottle, the long lunches — and threading those words back into your descriptions, your menu, your responses, your bio.

This is not keyword stuffing. The machine has gotten better than that. It is occasion-matching. AI recommendation engines weight the language of intent very heavily, because intent is what the guest typed. A restaurant whose digital surfaces mention celebrations and private dining and special occasions gets matched to the anniversary query. A restaurant that only mentions modern American cuisine does not — even if it would have been the better choice.

Sit down for ten minutes. List the top five reasons a guest comes through your door. Then search your own website for those exact phrases. Count how many appear.

If the number is low, you are not being skipped because you are wrong for the moment. You are being skipped because the machine has no way of knowing you are right for it.

— Margaret

Dear Operator,

A restaurant that tries to be everything to everyone becomes invisible to everyone in machine-mediated discovery. I have watched this happen so many times I now treat it as a rule.

When a guest asks an AI system where can I take a group of eight for a birthday dinner, the model has to make a confident decision in roughly four seconds. It needs to know which restaurants welcome groups, celebrate occasions, and can hold that energy without breaking. If your digital presence does not signal those things — clearly, plainly, in the language a machine can quote — you are invisible to the query. Even if you would have been perfect for it.

Audience signal is the work of being legible about who you are for. Price point. Occasion suitability. Group size. Formality. Dietary range. Accessibility. The hour of the day you come alive.

The most common failure I see is not the wrong signal. It is the absent one. Silence, to a machine, is not neutrality. Silence is irrelevance.

Try this exercise. If a model had to describe your ideal guest in one sentence using only what is visible in your digital presence, what sentence would it write? Now compare that to the guest you actually want.

If the two sentences are different, you are not failing your guests. You are failing the system that decides whether they hear about you at all.

— Margaret

Dear Operator,

Your regulars are your strongest signal — and most of them are invisible to the machine.

AI recommendation systems do not only read what you write about yourself. They read behavior. Reviews that mention we come here every week. Check-in frequency on Google Maps. Saved-to-collection patterns on Yelp. Tagged posts on Instagram from the same handles, month after month. These are loyalty traces, and the machine reads them as quality confirmation. If people keep coming back, the recommendation confidence rises.

But the trace has to be visible. A regular who never reviews, never checks in, never tags a photo, never writes the words our spot — that loyalty exists in your dining room and nowhere else. The machine cannot see it.

Loyalty architecture is not a punch card. It is the digital footprint of repeat behavior. It is asking your most loyal guests to leave one sentence in a review that contains the word again. It is making sure the language of return — every week, our place, where we always go — is somewhere in your reviews, somewhere in your tagged posts, somewhere a machine can find it.

Pull your last twenty reviews. Count how many mention return visits or regularity. If the number is below three, your loyalty signal is invisible — and the machine is treating you like a stranger to your own regulars.

You have built the room. Now make sure the room is legible from the outside.

— Margaret