Paid advertising for restaurants: the numbers that actually decide whether you keep spending

The ceiling on what you can pay for a new guest

Your maximum cost per incremental order equals that order's contribution margin, and with a 24 USD average ticket, 30% food cost and 2% for packaging and payment processing, the ceiling is 16.32 USD. Not a cent more, because payroll, rent and utilities never load onto the plate: they live in the month's break-even, and mixing them collapses the math into a number that looks prudent and is simply wrong. Now bring in the other side of the scale. The National Restaurant Association put sector operating margin between 3% and 5% for 2025, so an operation billing 1.2 million a year keeps 36,000 to 60,000 USD of clean profit. Fifteen thousand dollars of badly measured advertising per year eats between 25% and 41% of that figure, and nobody notices until the fiscal close. Platform ROAS measures correlation wearing a causation label, which is how a reported 5.2x coexists with zero additional dollars in the till.

Why does platform ROAS lie without lying?

Meta credits as a conversion any order placed within the seven-day window after a click, including the one from your Tuesday regular who was coming anyway.

A restaurant in Guadalajara spent 4,800 USD across three months of reach campaigns, went from 3,100 to 9,400 followers and billed the same as the previous quarter; the dashboard and the P&L were both telling the truth. The test that settles the argument takes fourteen days: switch the campaign off in a comparable postal code and measure the gap. Where operators have run it with discipline, between 45% and 70% of reported ROAS evaporates, and what remains is your real number. Almost the entire sector already sits where you want to advertise: 99% of restaurants keep at least one social media profile and 78% use Instagram, according to Restroworks 2025. That doesn't make advertising pointless; it means presence stopped being an advantage years ago and now only buys the right to compete for the same attention.

The ground you compete on: 99% of restaurants have a social profile

The operational consequence is awkward for anyone selling creative work. If 78% of your competitors post into the same feed, the differential comes from the asset behind the ad rather than the ad itself, and there Google rules: BrightLocal 2025 measured that businesses in the local pack top-3 hold 47 more reviews on average than positions 4 through 10. Forty-seven reviews cost less than 4,800 USD of paid reach and they don't switch off when you stop paying. Before raising the ad budget, look at what you already own: restaurant email marketing returns 36 USD per dollar invested according to Stripo 2025, while influencer marketing sits at 5.78 USD per dollar according to Socially Powerful 2025 and 7.65 USD per dollar in the iQFluence 2026 measurement, with average conversion of 2.55%. A sixfold gap, at minimum. The consultant's reading isn't «stop advertising».

Owned channels convert better than rented ones

It's that sequence matters: owned channel first, rented channel second, because a mailing list of 4,000 addresses is still yours when the algorithm changes, and a Meta audience is not. More than 90% of restaurants already run some rewards program (Paytronix 2025), so the contact data is sitting there, unexploited, in 90% of the operations that call me about paid advertising. A 12 USD CPA is ruinous if the guest comes once and cheap if they come six times, and that factor belongs to retention, not to the campaign. Paytronix measured in its Annual Loyalty Report 2024 a monthly member retention of 62% among the best QSR programs and 57.8% in full service. At 57.8% monthly, out of every 100 guests you acquire, 33 remain by month three and 19 by month six; lifetime value gets calculated on those, never on the 100 the dashboard shows.

Retention: the multiplier that decides whether your CPA is expensive

I apply it this way with my clients at Masterestaurant: paying 12 USD for a guest who leaves 16.32 USD of contribution per visit only works if the second visit exists. If your repeat rate isn't measured, you aren't buying customers, you're renting transactions at customer prices. Prepared food delivery in the United States moved roughly 96 billion dollars in 2024 according to Statista, and that figure is precisely why so many owners advertise inside the apps without doing the subtraction first. On a 24 USD order, platform commission takes between 15% and 30%, meaning 3.60 to 7.20 USD out of the same pocket that funds your CPA. Rerun the first passage's math with that bite included and your 16.32 USD ceiling drops to a band of 9.12 to 12.72 USD. There sits the paradox worth resolving: the channel delivering the most volume leaves the least margin available to buy that volume.

The delivery channel and the volume trap

You resolve it with destination, not budget — advertise toward your own ordering channel and let delivery handle discovery instead of scale. Translate the benchmarks into three scenarios before approving a single dollar. Small operation, up to 600,000 USD a year: at a 4% margin you keep 24,000 USD, a sensible ad budget runs 500 to 900 USD monthly, and the only metric worth demanding is cost per incremental order measured with a geographic switch-off. Mid-size operation, 1.2 to 2 million: 48,000 to 80,000 USD of profit, between 1,500 and 3,000 USD monthly, and it's time to reserve 20% of spend for switch-off tests in two postal codes. Group of four locations or more: consolidate reviews before scaling ads, because the 47-review gap of the top-3 (BrightLocal 2025) gets earned once and pays across all four sites, while CPA must be paid every month in each one.

Where these benchmarks come from and what they don't tell you?

The figures in this document come from verifiable public sources:

the National Restaurant Association for operating margins, Restroworks 2025 for social penetration, BrightLocal 2025 for the local pack, Paytronix for loyalty retention, Stripo and Socially Powerful for return by channel, Statista for the size of US delivery. None is proprietary research and none has been adjusted. The limits deserve saying out loud. Almost all of that measurement is US market, and commissions, tickets and labor costs across Latin America or Spain move on another scale; retention averages come from operations with a mature loyalty program, which is not the average case; and platform ROAS doesn't appear here as a reference because it isn't comparable across accounts. Use these numbers as a decision frame and measure your own in fourteen days. Platform ROAS is not a measure of causation, it is labelled correlation. Meta credits as a conversion any order falling inside the seven-day window after a click, including the Tuesday regular who was coming regardless.

Where the arithmetic of paid advertising breaks?

The way to size that is to switch the campaign off in a comparable postcode for fourteen days and measure the gap: where operators ran it, 45% to 70% of the reported ROAS disappeared.

Contribution margin outranks the check. A 24 USD order with 30% food cost and 2% of variable packaging and gateway cost leaves 16,32 USD of contribution, and that is your CPA ceiling. Payroll, rent and utilities do not load onto that order —they belong to the monthly break-even— and mixing them in drags the ceiling down to an unreal 4 or 5 USD that no campaign can ever meet. Guest lifetime value changes the scale of the problem. At 4,2 visits a year and a 24 USD average check, twelve-month accumulated contribution lands near 187 USD; paying 30 USD to acquire that guest is excellent business and paying 9 USD for someone who never returns is a disaster.

Where the arithmetic of paid advertising breaks — in practice?

The most repeated error is optimising the cost of the first purchase while ignoring the second entirely. Delivery conversion is decided before the ad runs.

A 24 USD aggregator order carrying a 28% commission gives up 6,72 USD of contribution, so your CPA ceiling drops from 16,32 to 12,90 USD or lower; the same ad that was profitable on the owned channel stops being profitable on the aggregator without a single pixel changing. Online reputation is a gate, not a parallel channel. Kim Ellis, research director at the National Restaurant Association, has argued publicly that the public score remains the most decisive filter for a guest who does not know the venue, and review-platform data across the sector shows that moving from 3,9 to 4,5 stars shifts revenue by 5% to 9%. Buying traffic into a 3,8-star listing is paying for more people to see the problem.

Where the arithmetic of paid advertising breaks — key points?

The decision calendar matters as much as the creative. Thirty days of waiting at a 26 USD CPA burns an extra 1.400 USD on a modest budget;

a fourteen-day kill and reallocation recovers that money for the campaign that does convert, and algorithmic learning does not break if you hold the ad set and move budget between creatives.