How Much Should a Restaurant Spend on Marketing? Build the Budget Backward From the Check
Build a restaurant marketing budget from conservative contribution per completed first visit—not a universal percentage or daily ad spend.
Your restaurant should spend only what a completed first visit can repay: set the maximum total acquisition cost below the conservative contribution margin from that first check, then multiply it by the number of completed first visits needed for a useful decision. Do not start with a percentage of revenue or somebody else's daily budget; translate your total test budget into a daily cap only after you choose the visit target and decision window.
Restaurant advertising examples that tracked visits are useful because they reach the result that matters. Responses, claims, and reservations can tell you where interest stopped, but they do not pay the check.
“One of the things that I'm trying to learn is how to calculate my cost per consumer.”
Source: An independent pizzeria owner
1. Decide what counts as a completed first visit
Choose one paid-marketing outcome before you spend: an identified first-time guest arrives, and the visit is connected to a settled check. Write down the exact records that prove each part so a response or reservation cannot quietly become a “customer” in the report.
Use these definitions for the test:
| Record | Count it when | Do not call it |
|---|---|---|
| Response | A person completes the campaign's trackable action | A lead unless you captured a permitted identifier |
| Identified lead | A unique person supplies the identifier required by your tracking process | A reservation or visit |
| Claim or reservation | The identified person accepts the offer or books a time | A completed visit |
| Completed first visit | That identified person is recorded as arrived or redeemed and was not already in your chosen lookback period | A matched check unless the POS record is connected |
| Matched settled check | The completed visit connects to a paid POS check under your written matching rule | An incremental sale caused by the ad |
Also record covers separately. One completed visit may bring several diners, while one diner may return later. Mixing those units makes both the cost per visit and repeat rate unreliable.
This step leaves you with one counting rule: completed first visits are the denominator for the acquisition budget; matched settled checks supply the first-visit sales. If your current setup cannot connect those records, fix that before using check economics to raise spend.
2. Subtract the costs from the first check before setting an ad limit
Calculate what the first incremental check contributes after the costs that rise because the visit happened. Use a conservative check value and current store costs, not the highest check from a Friday night.
Start with settled food-and-beverage sales under one written POS definition. Exclude pass-through taxes and tips, subtract refunds, and state whether the sales field is before or after discounts. Then subtract:
- food and beverage cost for the items sold;
- incremental labor, such as an added shift or overtime caused by the volume;
- discounts and rewards not already reflected in the settled-sales field;
- payment, ordering, or delivery fees that rise with the check; and
- any other fulfillment cost that occurs only because you acquired that visit.
The result is the conservative contribution margin per first visit:
Conservative settled sales − food and beverage cost − incremental labor − discounts or rewards not already reflected in sales − variable processing or platform fees − other visit-level fulfillment cost.
Do not subtract the same discount twice. If settled sales are already net of a discount, the lower sales figure already carries that cost. If the POS field is before discounts, record the discount as its own line.
Rent, salaried management, and other costs that do not change during the test do not belong in this per-visit calculation. They still matter to the restaurant, but they do not tell you how much another first visit can contribute toward the ad and production bill.
3. Set the maximum acquisition cost below the first-visit margin
Your maximum total acquisition cost per completed first visit must sit below the conservative contribution margin from that visit. The gap is your safety room for unmatched checks, forecast error, and a night that does not look like the average.
Write the cap in dollars before launch:
Maximum total acquisition cost per completed first visit < conservative contribution margin per first visit.
Total acquisition cost includes media plus the campaign's production or food-influencer cost and its share of software, management, or agency fees. It does not include food, discounts, or visit-level fees already subtracted from first-visit contribution.
Do not add assumed future visits to make the first visit look affordable. Measure repeat behavior later with a fixed window:
Repeat-visit rate = the subset of unique eligible identified first-time guests in the denominator who completed another verified visit inside the fixed window ÷ unique eligible identified first-time guests whose full fixed return window elapsed and whose return outcome was reconciled and known. Hold early returns out of the rate until cohort maturity, and report open, immature, ineligible, and unresolved counts separately.
Repeat visits may justify more spending after you observe them. Until then, the first check has to carry the acquisition decision on its own.
4. Turn the completed-visit target into one test budget
Choose the smallest completed-first-visit count that would change your decision, then make sure losing the full test budget would not threaten payroll or a required bill. If you would dismiss the result as “too few visits to tell,” the test is too small before it starts.
Write four items on the campaign brief:
- the number of completed first visits you need before making the decision;
- the maximum total acquisition cost you will accept for each one;
- the final date when every claim or reservation in the test will have a known outcome; and
- the maximum cash you permit the test to spend.
Now calculate the available media money:
Total test acquisition budget = target completed first visits × maximum total acquisition cost per completed first visit.
Media test budget = total test acquisition budget − test-specific production, food-influencer, software, management, and agency cost.
Daily media cap = media test budget ÷ the number of days in your chosen test window.
If the media result is zero or negative, the planned production and management cost already consumes what the first visits can support. Change the offer economics, lower the non-media cost, choose a different goal, or do not launch. Do not hide the problem by assigning the whole budget to “ad spend.”
One recorded Feast planning conversation used a hard cap of $20 a day for 14 days, which is a $280 media limit. That was one restaurant's test plan, not proof that $20 a day or a two-week window is enough for yours.
“There should be a settings dashboard where you put a maximum spend for any of the testing to make sure it doesn't go AI crazy.”
Source: A restaurant owner discussing a campaign test
A fixed decision date can help you avoid paying indefinitely, but the calendar is not the verdict. A day-14 review is one possible prechosen checkpoint, not a universal rule for every restaurant, audience, or offer.
5. Give one audience and one offer enough money to answer one question
Put the initial media budget behind one audience, one offer, and one primary creative comparison. That keeps the visit economics constant while you learn which creative earns more qualified responses.
Changing four offers at once creates four different check values, discount costs, response rates, and staff instructions. A Feast account review found one campaign budget divided among four offers at $10 a day each; the split left too little clean evidence to identify which offer deserved more money.
Keep the restaurant and offer recognizable before presenting the incentive. Putting the offer last lets your creative comparison test how well the restaurant earns attention without turning the whole ad into a coupon card.
If one path gets a few reservations and another gets a few replies, you still cannot tell which produces completed visits at an acceptable cost. Concentrating the budget gives the same audience and offer enough chances to reach the check-level result you chose in step one.
6. Keep every marketing cost on its own line
A monthly marketing budget needs separate lines for money paid to reach people, money paid to run the work, and costs created when guests visit. Combining them makes a cheap media result look profitable even when the all-in acquisition cost is not.
| Budget line | Put this here | How it enters the decision |
|---|---|---|
| Media | Dollars charged by the ad platform | Part of total acquisition cost |
| Software or management | Subscription, agency, or service fee allocated to this campaign | Part of total acquisition cost; state the allocation rule |
| Production or food influencers | Filming, editing, food-influencer fees, food credits used to produce the ad, and usage permissions | Part of total acquisition cost; separate reusable production when appropriate |
| Discounts or rewards | Price reduction or promised reward redeemed on a completed visit | Subtract from sales once, either in net sales or as its own visit-level cost |
| Restaurant fulfillment | Food, incremental labor, payment fees, delivery fees, and other costs caused by the visit | Subtract when calculating first-visit contribution |
| Fixed restaurant overhead | Rent, existing salaried labor, insurance, and costs unchanged by the test | Keep visible in the restaurant budget, but do not pretend one extra visit created the full monthly bill |
If one invoice includes several lines, split them for the decision or note what the package includes; never add the same charge twice. This separation lets you answer both questions: “How much cash will leave this month?” and “Can another completed first visit contribute enough to repay its acquisition cost?”
A revenue-percentage check can warn that the plan is unaffordable for your cash position, but it cannot replace those check-level answers.
7. Use completed visits and settled checks to continue, repair, or stop
Make the decision from one scorecard after the claims and reservations have had time to become known outcomes. Daily delivery, views, replies, and reservation counts help find a broken step; they do not overrule completed-visit cost or check contribution.
| Scorecard field | Exact definition |
|---|---|
| Total acquisition cost | Media plus production or food-influencer cost plus allocated software, management, and agency cost |
| Responses | Completed campaign actions under one stated definition |
| Identified leads | Unique people with the permitted identifier required by the campaign |
| Claims or reservations with known outcomes | Past-dated records marked completed, no-show, or canceled; exclude pending, future, and rescheduled records |
| Completed first visits | Identified first-time guests recorded as arrived or redeemed |
| Matched settled checks | Settled POS checks connected to those completed first visits |
| Directly matched settled sales | Sales from matched checks under the POS definition written before launch |
| Claim/reservation-to-visit rate | Completed first visits ÷ claims or reservations with known outcomes |
| Total acquisition cost per completed first visit | Total acquisition cost ÷ completed first visits |
| Matched first-visit contribution | Directly matched settled sales minus the recorded visit-level costs for those checks |
| Check-match coverage | Completed first visits with a matched settled check ÷ all completed first visits |
| Repeat-visit rate | The subset of unique eligible identified first-time guests in the denominator who completed another verified visit inside the fixed window ÷ unique eligible identified first-time guests whose full fixed return window elapsed and whose return outcome was reconciled and known; hold early returns out until cohort maturity and report open, immature, ineligible, and unresolved counts separately |
Directly matched settled sales show a recorded connection; they do not prove every matched sale was caused by the ad. Do not turn attributed gross sales into incremental profit. When you have a defensible comparison, calculate conservatively estimated campaign contribution as:
Conservatively estimated incremental completed first visits × conservative contribution margin per first visit − total acquisition cost.
Write the decision rules before launch:
- Continue when the tracking is intact, enough outcomes have matured for your written decision, estimated incremental campaign contribution is positive after all included costs, and the restaurant has room to serve more guests.
- Repair when the first broken step is clear and the test remains inside its cash cap. Fix tracking before creative, response before reservations, attendance before check value, and check matching before any profit claim. If delivery was healthy and then abruptly stopped, use the cliff diagnosis before replacing everything.
- Stop when the cash cap or decision date arrives and the records still cannot support a decision, or when conservative contribution remains below acquisition cost. Preserve the result so the next test does not repeat the same failure.
If you lack a defensible estimate of incremental visits, report matched sales, contribution, and acquisition cost separately. That is enough to identify the next repair; it is not enough to claim causal profit.
Feast connects the ad response to the settled check
Feast becomes relevant when the budget depends on completed first visits and matched checks but those records live in separate systems. Its published Full Package captures a phone identity through a campaign page, matches later in-person purchases through supported POS connections, and reports POS-linked sales by campaign.
As captured September 1, 2026, Feast lists that package at $750 per month per location plus the restaurant's media budget, month to month. The published package includes food-influencer sourcing and compensation, editing, ad management, and POS-linked reporting; do not add those included services a second time.
Feast can reduce the handoffs between the ad response and the check. Its POS-linked campaign sales remain attributed sales, not automatic proof of incremental profit or causal lift; your restaurant still supplies its costs, capacity, and comparison.