How Do I Know If My Restaurant Ads Are Working?
See which evidence shows whether your restaurant ads are working, where a campaign is breaking down, and whether the sales leave any profit.
Your restaurant ads are working when they create enough additional completed visits and sales to leave positive contribution after campaign costs. To find out, ask where guests disappear: did enough people see the ad, respond, arrive, and pay?
This page is for diagnosing an ad that is already running. If the restaurant cannot yet connect reservations, offer claims, or awareness campaigns to its records, set up restaurant ad tracking first. Then use the first weak step to decide whether to keep spending, repair the campaign, or stop it.
“Every month I can deliver stats like clicks from Meta and Google ads for my clients, but it doesn't tell them the full story of how that turns into revenue.”A restaurant marketer working with independent restaurants
Decide what working was supposed to mean
Ask what the campaign was supposed to change inside the restaurant. A reservation or offer campaign should produce visits and paid checks, while an awareness campaign should create a credible increase in sales, covers, transactions, or sales of the item being promoted.
Then compare the result with the complete campaign cost. Include the ad spend and any applicable creative, production, discount, food, incremental labor, processing, software, or management costs. The ad is economically working only when the additional sales leave enough contribution to repay those costs.
Before blaming the ad, make sure the result is not disappearing from the records
Before calling an ad weak, confirm that the restaurant can still see what happens after the click. Open several recent responses and follow them through the records you already have.
For a reservation campaign, check whether the source stayed attached when the guest booked, arrived, and paid. For an offer campaign, check whether the claim appears beside the redemption and settled check. For an awareness campaign, confirm that the POS periods being compared use the same locations, days, hours, and sales fields.
If responses are arriving but the later records are blank, you do not yet know that the campaign failed. You know that the connection failed. Cash payments, a companion paying, changed contact information, missed staff steps, cancellations, and no-shows should remain visible as unmatched outcomes rather than being silently counted as sales or ignored.
Feast can capture a guest through an offer, connect a later purchase through supported POS data when the identity matches, and report those sales by campaign. Even that path can miss a real visit when another person pays or the restaurant does not preserve the identifier.
Find the first place the campaign loses people
Once you trust the records, the pattern tells you where the problem is.
| What you see | What it supports | What to investigate next |
|---|---|---|
| The ad barely delivers | The restaurant is not buying enough attention to judge the offer | Rejection notices, billing, dates, budget, location, and audience eligibility |
| The ad delivers but few people respond | People see it but do not take the intended action | The opening creative, the reason to visit, the offer, and whether the next step is obvious |
| People respond but few arrive | Interest is not becoming restaurant traffic | Booking friction, reminder timing, offer availability, hours, address, and no-shows |
| Guests arrive but few checks match | Visits are happening but the sales connection is weak | Staff recognition, payer identity, redemption handling, and the POS handoff |
| Checks match but contribution is negative | The campaign creates recorded sales without repaying its costs | Check size, discount depth, food and labor cost, fees, and ad spend |
| Direct matches look modest but restaurant sales move | The campaign may be affecting people who never identified themselves | Compare covers, transactions, item sales, and like-for-like revenue before estimating the wider effect |
Repair the first broken part. If the ad is not delivering, rewriting the offer does not solve the immediate problem. If people are responding but not arriving, buying more clicks only sends more people into the same leak. If matched sales are healthy but contribution is negative, the problem is the economics rather than the creative.
Awareness ads require a wider restaurant comparison
Awareness ads are harder to judge because most viewers will not identify themselves before visiting. A person can watch a food video, remember the restaurant, visit later, and pay without leaving a direct campaign trail.
Compare the campaign period with a relevant baseline using the same locations, weekdays, hours, and POS definitions. Look at settled sales, covers or transactions, average spend, and sales of the promoted dish or daypart. Record holidays, weather disruptions, closures, local events, price changes, promotions, or service problems that could explain the movement. The result is observed change until a stronger comparison supports an incremental estimate, and the word “estimated” should remain attached to it.
Keep spending, repair, or stop
Once enough visits and checks have had time to happen, put three numbers beside one another:
- the complete campaign cost;
- conservatively estimated incremental sales; and
- estimated contribution after campaign costs.
Do not treat attributed gross revenue as cash available to spend again. Food, labor, discounts, payment fees, fulfillment, creative, management, software, and advertising all consume part of the sale. A campaign can report a strong ROAS and still lose money if the credited sales were not incremental or the cost list was incomplete.
Keep spending when the tracking is intact, a conservative estimate shows positive incremental contribution after campaign costs, and the restaurant can serve more demand.
Repair the campaign when the symptom table shows a specific break between delivery, response, arrival, payment, or contribution. Change that step first and keep the rest of the campaign stable enough to learn whether the repair worked.
Stop after the agreed test budget or time window when the strongest defensible result remains below break-even. Do not keep spending only because the platform still reports clicks, reservations, or attributed revenue.
If the campaign is profitable on a small budget, do not assume the same return will survive a large increase. Restaurant ROAS often changes as ad spend grows, so judge the contribution produced by the added dollars.
Feast shows the direct sales and the wider restaurant result
Feast helps restaurants preserve direct campaign-to-check connections when the guest and supported POS record can be matched. It also lets the owner examine those campaign results beside wider restaurant sales and visit movement instead of pretending every response was a purchase.
Feast Revenue Analyzer helps you examine how your campaigns relate to restaurant sales, then use those findings to plan what to test or invest in next. Scale your marketing with confidence and iterate with clarity. Analyze my restaurant's revenue.