How to tell if marketing brought in your walk-ins

Partly: some walk-ins can be tied to the marketing behind them, and the rest can at least be counted. Here is what that looks like, starting at the host stand at 7:30 on a Friday.

Picture eleven people waiting by the door. A couple near the front saw your Reel on Thursday, talked about the new menu over dinner at home, and walked in without booking. The host seats them in four minutes. They spend $140, pay at the table and leave happy. Everything about that evening worked except the record of it: as far as any report can tell, two strangers appeared and paid.

That is the walk-in problem. The sale is real and the register has it. What's missing is any link between the check and whatever the guest saw earlier, and a report can't invent that link afterward without guessing. The honest move is to find the places where a link can be made without slowing the room, and then to track how often it happens.

What does capture look like on a full Friday?

The measurement improvements that last happen in the room, not in the ad account, and they have to survive the worst hour of the week. A step that only works when the restaurant is quiet isn't a process. Three that tend to hold up:

• The wait. A guest who gives a name and number for a table has already identified themselves. The host stand is the one place on a full Friday where a guest stands still and expects to be asked something.
• The check. Guests are waiting for the check anyway, which makes it a natural moment for an optional step: a code on the presenter that offers something worth having, like a dessert on the next visit or early word on the next menu.
• The bar. Bar guests are often the walk-ins with the most time and the least reason to book. A card on the bar that makes the same offer reaches them while they wait for the next round.

Whatever you choose, the guest should understand why they're sharing it and get something useful back, and a server should be able to offer it in one sentence while carrying plates.

How much of your room can you actually see?

Here is a made-up Friday. Say a 60-seat room seats 90 parties:

• 34 booked ahead, so a reservation sits behind each of them.
• 12 gave a name and number for a wait.
• 9 scanned the code on the check and took the offer.

That is 55 of 90, or 61%, who left something a report can use. The other 35 are walk-ins as far as any report is concerned. They may have seen your ads. They may have just walked past. Nobody can say.

Now take the same room on a Tuesday: 40 parties, 14 of them identified, or 35%. Same restaurant, same marketing, but Friday's report can see nearly two thirds of the room and Tuesday's can see about a third.

That share is the number worth tracking. It tells you how much of the room your reporting can currently see, and it makes every other number more honest. If Tuesday's campaign looks quiet, part of the reason may be the report, not the campaign.

What do you do about the guests you can't see?

Leave them unassigned. A walk-in with no signal belongs in the broader picture of the business, but it shouldn't be credited to a campaign as a verified sale. Crediting them anyway is how a report ends up with a number nobody can defend, which is the same problem behind why Google and Meta ROAS don't match your register.

The restauWant app works the same way. It ties what guests see in the room to the campaign it belongs to, and walk-ins with no proof stay out of the verified number instead of being guessed at. You can see how that fits with the rest of your marketing on the in-venue page.

You don't need us to start. This week, pick one Friday and one Tuesday. Count the parties, count how many left something (a booking, a name for the wait, a scanned code), and write down both shares. You'll have your first honest number, and you'll know which night to improve first.

Next step: Use the restaurant marketing data setup checklist to map which of your systems already holds each part of the guest journey.