Most of Your Dinner Rush Is Decided on a Phone Twenty Minutes Earlier

No, and be careful with anyone who does. Map pack position is decided by Google on signals we can influence but never control: how close the searcher is standing, how relevant your profile and website are to what they typed, and how prominent your restaurant is in reviews, citations and coverage. Two people on opposite corners of the same street see different results, so there is no single position to promise anybody. What we can do is make the things you actually own correct and complete — categories, hours, menu, service attributes, photos, the pages on your site that match real searches, and a steady habit of asking happy guests for reviews. Then we measure your ranking across a grid of points around the restaurant rather than from one address, so movement is visible honestly. If a top position in your area is not realistic yet, we will say so before you spend money on it.

We can improve the odds and measure the result, but nobody controls what a model says. AI assistants assemble answers from your site, your Google Business Profile, review platforms, local press and directory listings, and they change what they say from week to week with no notice and no appeal. The work is unglamorous: make your cuisine, neighbourhood, price band, hours and standout dishes stated plainly in text a machine can read, keep your profile and third-party listings consistent, and earn the local mentions those systems tend to draw on. Then we run the questions your guests would actually ask across several assistants each month and record whether you appear and who appears beside you. That gives you a trend line instead of a guess. What it does not give you is a guarantee, and any agency offering one is selling something it cannot deliver.

Sometimes, and often not for the obvious searches. Bidding on generic terms like dinner near me puts you against every restaurant in the area for traffic that converts unpredictably and at a thin margin per cover. Where paid search tends to earn its place for restaurants is the higher-value work: catering, private dining, buyouts, holiday parties and large-party bookings, where one booking can be worth a month of ordinary ad spend. We would rather start there, prove the cost per booked event, and expand only if the numbers hold. Before any of that, we check that your enquiry form, phone and booking links are actually tracked, because spending on ads without knowing which enquiries came from them is how restaurants conclude that advertising does not work. If the arithmetic does not support ads for you, we will tell you that instead.

No. Your reservation system handles tables, covers and the floor plan, and it does that better than anything we would build. What it does not do is bring people to the point of booking or catch the enquiries that arrive outside it — the catering email, the Instagram message about a fiftieth birthday, the voicemail left during service. Our work sits on either side of the reservation platform rather than on top of it. Search visibility and ads bring the demand in; the enquiry response layer makes sure the ones that do not arrive as a neat online reservation still get answered quickly and captured somewhere you can see. If your platform already has a feature that covers part of this, we will use it rather than duplicate it, and we will say so rather than quietly bill for something you own.

The parts move at very different speeds and it is worth separating them. Enquiry response is immediate — if event emails currently wait a day, answering them within minutes changes what happens to next week's enquiries, and you can read the transcripts yourself in the first week. Google Business Profile fixes, categories, photos and menu accuracy often show movement within a few weeks because you are correcting things that were simply wrong. Paid search gives you a readable cost per booked event within a month or two, once tracking is clean and there is enough volume to judge. Local SEO and AI visibility are the slow ones, and we would not draw conclusions inside a quarter. Anyone promising you map pack results in thirty days is guessing, and we would rather set the expectation correctly now than explain a miss later.