A restaurant KPI dashboard needs exactly 8 numbers, tracked weekly on one page: sales, covers, average check, food cost %, labor cost %, prime cost % (food + labor, target at or below 60%), food cost variance (actual vs theoretical, target under 2 points), and progress against break-even. Everything else is commentary. The discipline is the weekly cadence, not the dashboard software.
When I was responsible for 14 locations and a $10-12M P&L, the tool that held it together was not an enterprise BI platform. It was one page per site, 8 numbers, every Monday morning, ranked best to worst. The format barely matters. The cadence and the shortlist are everything. Here is the exact dashboard, why each number earns its place, and a free Excel template with the formulas already built.
Skip ahead if you want the spreadsheet - 13 weeks of formulas, plus a definitions sheet with targets.
Download the free KPI template (.xlsx)Why Weekly, and Why Only 8 Numbers
Monthly reporting is too slow for a business that buys perishable inventory twice a week and reschedules labor daily - by the time a monthly P&L flags a problem, you have paid for it 4 times. And 30-metric dashboards fail in the other direction: when everything is highlighted, nothing is. Eight numbers fit on one page, take 20 minutes to review, and cover the 3 questions that decide survival: did we sell enough, did we keep enough of it, and are we above the line that pays the bills?
The 8 KPIs That Earn Their Place
| KPI | How to calculate it | Target / healthy range |
|---|---|---|
| Sales | POS total for the week | Vs same week last year, not vs hope |
| Covers | Guests served | Trend - falling covers hide behind rising prices |
| Average check | Sales / covers | Track the trend; price moves show here |
| Food cost % | COGS (open inv + purchases - close inv) / food sales | 28-35% full service |
| Labor cost % | Fully loaded labor / total sales | 25-35% by format |
| Prime cost % | Food % + labor % | At or below 60% |
| Food cost variance | Actual % - theoretical (recipe) % | Under 2 points |
| Break-even progress | Week sales vs weekly break-even sales | Above 100%, known cold |
Notice what is not on the list: social followers, review scores, loyalty signups. Those matter, but they are diagnostics you pull when a core number moves. The dashboard's job is to tell you whether the operation made money last week and exactly which lever slipped if it did not - sales (covers x check), cost (food + labor), or structure (break-even).
Prime Cost Is the Headline Number
If you only manage one row, manage prime cost %. Food plus loaded labor is the largest controllable share of every sales dollar, and the 60% line is where full-service operations separate profit from loss - on margins that typically net just 2-6% of sales, 5 points of prime cost drift is usually the whole year's profit. The dashboard template calculates it automatically from the food and labor inputs and shows the variance to 60% in its own column, because the gap is the action item.
Want the live version first? Put last week's numbers into the calculator and see your distance to 60%.
Run the prime cost calculatorHow to Run the Monday Review in 20 Minutes
- Fill the row for last week: sales, covers, food $ and labor $ are the only inputs - the rest calculates
- Read prime cost % first; if it is at or below target and variance is under 2 points, the review is nearly done
- For any number more than 2 points off target, name one cause - receiving, portioning, overtime, comps - not five
- Assign one fix, one owner, one recheck date the following Monday
- Multi-site: rank locations by prime cost and put the bottom two on a 15-minute call - manage the spread, never the average
The trap to avoid is letting the review become a meeting. It is not a meeting. It is a 20-minute instrument check, and its only outputs are a ranked list and at most 2 assigned fixes. The week you skip it because things feel fine is - reliably - the week a 4-point leak starts compounding.
Spreadsheet, Software, or Both?
Start with the spreadsheet. It costs nothing, takes 10 minutes a week to fill, and builds the habit that any future software depends on. Move to automated reporting when the manual version starts failing for a structural reason - multiple locations re-keying the same numbers, data arriving too late for Monday, or owners wanting daily visibility. Software accelerates a working cadence; it cannot create one. That sequencing - habit first, automation second - is exactly how we approach reporting builds for clients.
The Bottom Line
A restaurant KPI dashboard is not a software purchase - it is a weekly discipline with 8 numbers on one page: sales, covers, average check, food %, labor %, prime cost %, variance, and break-even progress. Download the template, fill in last week, and read prime cost first. If you keep that 20-minute Monday appointment for 13 straight weeks, you will know your operation better than most owners ever know theirs.