Almost every restaurant tracks numbers—far fewer track the right ones, at a cadence that lets anybody act on them. Sales, covers, and food cost get reviewed monthly, usually after the accountant closes the period, by which point the month that produced them is gone. The restaurant KPIs that matter are the handful that move before the P&L does — the ones that tell an operator on a Tuesday that something is wrong, rather than confirming it on the fifteenth of the following month. Getting that set right is the starting point of our growth and revenue work, because no marketing decision is defensible without it.
What restaurant KPIs actually are
A KPI is not simply a number the business produces. It is a number that a specific person can influence, on a timeline short enough to matter, and that reliably moves ahead of profit. Total revenue fails that test — everyone watches it, nobody owns it, and it tells you nothing about why it moved. Average check per guest passes: a general manager can influence it this week through menu placement and service standards, and its direction shows up in margin long before the period closes.
This distinction is why dashboards proliferate, and decisions do not. A restaurant group we work with arrived with a forty-two metric weekly report. Nobody read past the first page, and the two numbers that would have flagged a failing location — declining repeat visit rate and a slipping labor-to-sales ratio at one site only — were on pages three and four, averaged across the group so the problem was invisible. Fewer metrics, segmented, beats more metrics aggregated, every time.
The operational numbers
Four operational KPIs carry most of the signal. Prime cost — combined cost of goods and total labor as a percentage of sales — is the single most reliable health indicator in the industry, and the one number worth calculating weekly rather than monthly. Full-service operators generally target the low-to-mid sixties; quick-service runs lower. What matters isn’t hitting an industry benchmark, but knowing your own trailing figure well enough to notice a two-point drift the week it happens.
Average check per guest, cover count by daypart, and table turn time fill in the rest. Read together, they explain almost any revenue movement: a soft week is either fewer people, less spend per person, or slower throughput, and each has a different fix. Read separately, they mislead. A rising average check alongside falling covers is not a win — it usually means the price-sensitive half of the guest base has quietly stopped coming, a problem that compounds and that menu pricing strategy either caused or can correct.
Two adjacent numbers deserve the same weekly attention. No-show rate on reservations is pure lost margin on covers already committed, and it is measurable and reducible — the mechanics of that sit in our guide to reducing restaurant no-shows. Labor efficiency by shift, rather than labor as a monthly percentage, is where operational efficiency gains are actually found, because the overstaffed shifts are never the ones anyone remembers.
The marketing and discovery numbers
Marketing KPIs are where restaurant reporting goes furthest wrong, because the available metrics are mostly the wrong ones. Follower counts, impressions, and reach are reported constantly and predict nothing. The marketing numbers that correlate with revenue are narrower: repeat visit rate, cost per acquired guest, review volume and velocity, and share of discovery — how often the restaurant appears when someone in the trade area searches for what it sells.
Repeat visit rate is the one most operators cannot produce and most need. A restaurant that retains thirty percent of first-time guests has a fundamentally different economic model from one retaining fifteen, and no amount of acquisition spend closes that gap. Review velocity — new reviews per month, not lifetime average rating — is the closest available proxy for whether recent guests are satisfied, and it feeds directly into local visibility, which is why review management is an operational discipline rather than a marketing chore.
Share of discovery is newer and increasingly decisive. It is no longer enough to rank in a map pack; a growing share of restaurant intent is resolved inside AI answers and assistant results that never produce a click. Measuring whether a restaurant appears in those answers belongs on the same dashboard as cover counts, and it’s why local SEO now has to be assessed by presence rather than position alone. Where a restaurant is absent from the conversations its competitors appear in, a content gap analysis is usually the fastest way to see it.
Cadence beats comprehensiveness
The most common failure we see is not tracking the wrong KPIs. It is tracking reasonable KPIs on the wrong clock. Prime cost reviewed monthly is a history lesson; reviewed weekly, it is a control. Repeat visit rate reviewed annually is trivia; reviewed quarterly, it changes where the budget goes. The discipline that separates operators who use their numbers from operators who merely collect them is a fixed weekly review of six or seven figures, owned by a named person, with last week beside this week.
That also determines how marketing money should be allocated. A KPI set that stops at revenue and food cost cannot tell you whether a campaign worked, so spend gets defended by anecdote. A set that includes cost per acquired guest and repeat visit rate makes the restaurant marketing budget an evidence-based argument, and makes retention programs such as loyalty measurable rather than assumed.
Frequently asked questions
What are the most important KPIs for a restaurant?
Prime cost, average check per guest, cover count by daypart, and table turn time cover most operational signals, and repeat visit rate, cost per acquired guest, and review velocity cover most marketing signals. Seven numbers reviewed weekly outperform forty reviewed monthly, because a KPI only counts if a named person can influence it before the period closes.
What is a good prime cost for a restaurant?
Full-service restaurants generally target a prime cost in the low-to-mid sixties as a percentage of sales, and quick-service concepts run lower. The benchmark matters less than the trend: knowing your own trailing figure well enough to notice a two-point drift in the week it happens is worth more than matching an industry average.
How often should restaurant KPIs be reviewed?
Review operational KPIs such as prime cost, covers, and labor efficiency weekly, with last week beside this week. Marketing KPIs such as repeat visit rate and cost per acquired guest are better read quarterly, because they move slowly and weekly readings produce noise rather than decisions.
Are social media followers a useful restaurant KPI?
No. Follower counts, impressions, and reach are reported constantly and predict nothing about revenue. The marketing numbers that correlate with revenue are repeat visit rate, cost per acquired guest, review volume and velocity, and share of discovery — how often the restaurant appears when someone in the trade area searches for what it sells.
If your weekly numbers don’t show where margin and guests are going, request a growth diagnostic.