Five numbers to check every Monday: restaurant KPIs for owners with no time for spreadsheets

Picture the first Monday in February. December was a blur, January was quiet, and you cannot tell whether last week was bad or just normal.

Knowing your normal takes five numbers, checked every Monday on one page: sales against the same week last year, covers and average spend, prime cost as a share of sales, labour against sales, and cash against your 13-week forecast. Add staff turnover once a month. Each has a clear "look closer" line, so ten minutes on Monday tells you whether last week was normal. Last checked: 5 November 2026.

Why five numbers, and why on a Monday?

Because owners with no time do not read dashboards with forty tiles. Five numbers cover the questions that decide whether the business survives the year: are guests coming, are they spending, are the two biggest costs under control, and will the bank account cope with what is due?

Monday works because the week is complete, the weekend is in the numbers, and there is still time to change this week's rota and orders. Do it at the same time each week, with the same five figures, so the trend becomes obvious.

What the scorecard leaves out on purpose: social followers, review counts, table turns and dozens of item-level reports. They matter, but they explain a change in the five numbers; they do not tell you whether there is one.

KPI 1: are sales up or down on the same week last year?

Compare net sales (before VAT or sales tax) with the matching week last year, not with last week. Week to week, a restaurant swings with weather, paydays and school holidays; year on year, most of that cancels out.

• Match by weekday. Compare Monday 11 January 2027 to Sunday 17 January with Monday 12 January 2026 to Sunday 18 January, not with the same calendar dates, or one week will have an extra Saturday.
• Watch moving holidays. Easter Sunday is 5 April in 2026 and 28 March in 2027, and UK Mother's Day moves with it. Compare those weeks with each other, or note the shift beside the number.
• Use a three-week average to read the trend. One bad week is weather; three is a pattern.

Look closer when: sales are more than 5% below last year for three weeks in a row, or the gap is widening.

KPI 2: is the change from covers or from average spend?

Split sales into covers (guests served) and average spend per cover (net sales divided by covers). The same 5% fall means different things:

Count covers the same way every week. For counter service, use transactions instead of covers.

Look closer when: average spend falls more than 3% against last year while your prices have gone up, which means guests are trading down.

KPI 3: what is your prime cost this week?

Prime cost is food and drink cost plus all labour cost (wages, employer taxes, pension or benefits), as a share of net sales. It combines the two costs you can change fastest. Our profit margin guides for the UK and US explain the definition and give published benchmarks.

Weekly food cost is only accurate if you count stock weekly. If you do not, use purchases for the week as a stand-in and correct it at each month-end count.

Look closer when: prime cost moves more than two points from your own four-week average. A one-off is usually a delivery timing; two weeks running is a real change.

KPI 4: is labour in line with sales?

Labour as a share of sales tells you whether the rota matched the trade. Sales per labour hour (net sales divided by hours worked) tells you the same thing in a way staff understand, and it is not distorted by a pay rise.

Worked example. A cafe takes £9,600 ($12,800) net in a week with 480 hours on the rota: £20 ($26.67) per labour hour. In the first week of January, takings drop to £7,900 ($10,500) and the rota still has 470 hours: £16.81 ($22.34) per hour. To get back to £20 an hour on those takings, the rota needs about 395 hours, 75 fewer. That is the conversation for Monday's rota, not month end.

For UK wage rates and how to cost an hour of labour, see our labour cost guide. In the US, check local scheduling laws before cutting published shifts; see our guide to predictive scheduling laws.

Look closer when: sales per labour hour is more than 10% below your four-week average.

KPI 5: is the bank where your forecast said it would be?

Compare this Monday's bank balance with the balance your 13-week cash flow forecast expected. Profit can look fine while cash runs short, because of VAT or sales tax, supplier terms and quarterly bills. Our 13-week cash flow forecast shows how to build one.

Look closer when: the actual balance is more than one week of takings below the forecast, or any week in the next six falls below your buffer.

Monthly: how fast are staff leaving?

Staff turnover is the slow number that drives the fast ones. Every leaver costs recruitment time and a few weeks of a slower, less confident team member.

Monthly staff turnover = leavers this month / average headcount x 100

For a benchmark, the US has the only reliable public figure. In August 2026 the Bureau of Labor Statistics put the quits rate for accommodation and food services at 3.5%, the highest of any industry, against 1.9% for all jobs (BLS, 13 October 2026). The next release, on 3 November 2026, revised August to 3.6% and put September at 5.0% (preliminary) against 2.1% for all jobs (JOLTS).

Over a full year the gap is starker. In our analysis of the BLS Job Openings and Labor Turnover Survey (not seasonally adjusted monthly quits rates, downloaded 5 November 2026), accommodation and food services recorded quits equal to about 51 per 100 jobs in 2025, against about 24 per 100 across all jobs. The sector's average monthly rate was 4.2% in 2025, down from 5.8% in 2022 and 4.9% in 2019. Those figures count only people who chose to leave, not dismissals or seasonal lay-offs.

There is no comparable official UK series for hospitality quits, so compare yourself with your own last year.

Look closer when: monthly turnover is above 6% for two months running (half as high again as the sector's 2025 average for quits alone), or any one role (head chef, a supervisor) leaves twice in a year.

The Monday scorecard to copy

MONDAY SCORECARD                     Week ending: [Sun date]   Site: [ ]

                                This week   Same week LY   4-week avg   Flag?
1. Net sales (ex VAT/sales tax) [ ]         [ ]            [ ]          [ ]
   Change vs last year          [ ]%        3-week avg change: [ ]%
2. Covers (or transactions)     [ ]         [ ]            [ ]
   Average spend per cover      [ ]         [ ]            [ ]          [ ]
3. Prime cost % of sales        [ ]%                       [ ]%         [ ]
   (food and drink cost + all labour cost) / net sales
4. Labour % of sales            [ ]%                       [ ]%
   Sales per labour hour        [ ]                        [ ]          [ ]
5. Bank balance today           [ ]   Forecast said: [ ]   Gap: [ ]     [ ]
   Lowest week in next 6 weeks: [ ] on [date]   Buffer: [ ]

FLAG RULES
1. Sales more than 5% below last year for 3 weeks in a row
2. Spend per cover down more than 3% on last year after a price rise
3. Prime cost more than 2 points off the 4-week average
4. Sales per labour hour more than 10% below the 4-week average
5. Bank more than one week of takings below forecast

MONTHLY (first Monday)
Leavers [ ] / average headcount [ ] = turnover [ ]%   Flag above 6%, 2 months

THIS WEEK'S ONE ACTION: [ ]
Owner: [ ]   Check again on: [ ]

Worked example: one Monday in January

Here is a 70-cover restaurant's scorecard for the week ending Sunday 17 January 2027, in both currencies. The figures are illustrative.

The reading: guests are spending more but fewer are coming, and the rota has not shrunk with them, which is why prime cost jumped. The one action this week is to take about 50 hours out of the Monday to Wednesday rota and look at why covers are down: bookings, walk-ins or reviews. Our January marketing plan covers the second part.

Can TablePort handle this for you?

Most of the scorecard comes from TablePort without a spreadsheet.

1. The daily report gives each day's sales and costs with an AI-written summary of what went well and what needs attention.
2. A health score for each day, worked out by fixed rules, so a bad day stands out at a glance.
3. Margin recommendations name items to promote or review.
4. Group reporting puts every location in one view if you run more than one site.

The honest limit: the health score is rule-based and looks back; it is not a forecast, and the bank balance still comes from your bank. See restaurant management.