Google Performance Max 'Store Visits' for Restaurants: Are the Numbers Real?

Google Store Visits are a modeled estimate of total visits after eligible ad interactions. They are not a person-by-person record of diners, a count of settled checks, or proof that Performance Max caused every visit in the column.

That is why Google Ads store visits accuracy is the wrong question if it means, “Can I treat this number as sales?” The useful question is whether Google’s estimate moves in the same direction as your own restaurant records. That distinction belongs at the center of any honest restaurant marketing ROI report.

An agency told us it had lost restaurant clients after they stopped trusting Google’s estimated Store Visits. That does not prove the estimate was false. It shows what happens when a modeled platform number is presented as if it were a restaurant transaction.

The number is modeled because Google cannot observe every diner directly

Google says Store Visits uses aggregated statistics to estimate visits after an ad interaction, and eligibility depends on factors such as sufficient ad activity, enough foot traffic, and properly connected location information. Google may also use opt-in survey data to check the model’s quality; Google explains the modeling and eligibility requirements here.

That makes Store Visits useful directional evidence. It can tell you that Google detected a pattern consistent with more visits after ad exposure. It cannot tell you which diner arrived, whether that person ate, how much the table spent, or whether the visit would have happened without the ad.

There is a second trap hiding beside the visit count. Google lets advertisers assign a value to a Store Visit conversion. That value is an input used for reporting and bidding; it is not revenue pulled from your point of sale. Google’s reporting documentation distinguishes Store Visits from the advertiser-set Store Visit value.

If the report shows a large conversion value, ask where that value came from before anyone calls it restaurant sales.

Reconcile the estimate with what happened in the restaurant

Start with the business result you expected the campaign to change. For a restaurant, that normally means completed visits, covers, settled checks, or some combination of them. Compare the same locations and the same days before and during the campaign, and note anything obvious that also changed: hours, weather, holidays, large events, pricing, staffing, or another promotion.

Do not compare Google’s Store Visits directly with revenue and expect an exact match. One is a modeled count of visits after ad interactions. The other is money from every paid check, including regulars and guests who may never have seen the campaign. Use covers or completed reservations to check traffic; use net settled sales to check the value of that traffic.

The goal is not to force the two numbers to agree or to claim that covers validate Google's model. It is to find out whether the modeled direction is consistent with the business outcome and investigate when it is not.

Use a trackable action when you need proof closer to sales

If the campaign promotes a reservation or offer, carry the campaign source into the booking or claim, mark whether the guest actually arrived, and connect the visit to a paid check when the identity matches. That is stronger sales evidence than a modeled Store Visit, although it still misses anonymous walk-ins, cash payments, and checks paid by someone else. The full Facebook and Instagram sales-tracking guide explains those paths and the awareness alternative.

The account setup can change the estimate even when the restaurant does not

Google notes that Store Visits reporting can change because of customer behavior, seasonality, or events, but also because of account configuration. Business Profile connections, location assets, conversion settings, the conversion window, and location targeting can all affect what appears in the report. Google lists those diagnostic areas in its Store Visits reporting help.

Before blaming the campaign or trusting a sudden jump, confirm that the right locations are connected and that the reporting window matches the period you are comparing. If a store opened, closed, moved, or changed hours, record that too.

This is why a single Store Visits number should never make the decision by itself. It can be wrong for your business question even when Google calculated the metric exactly as documented.

Believe the direction only as far as your independent evidence supports it

Use Store Visits as directional evidence. Reconcile it with completed reservations, offer claims, covers, and settled checks. Use a like-for-like comparison to describe observed change, and use a proper control only when you need to claim incrementality.

Then calculate economics from restaurant records. Directly matched sales are revenue, not profit. Contribution subtracts the variable costs tied to those checks and the full campaign investment. A “good” ROAS depends on that break-even point, not on a universal platform benchmark. Here is how to calculate a restaurant-specific ROAS target.

The honest conclusion may be that Google’s estimate looks consistent with your business data. It may also be that the estimate moved while the dining room did not. Both answers are more useful than accepting or rejecting the metric on faith.

Want to understand whether Google's reported store visits line up with your restaurant's sales? Feast Revenue Analyzer brings your marketing and POS history together to investigate the results in the context of your restaurant's seasonal patterns. Use those findings to decide what deserves more investment. Analyze my restaurant's revenue.