The P&L that turns up after everyone has moved on
It is the 17th. The finance team has just sent round last month's site P&Ls. The ops director opens the pack and sees that one of the city sites ran its food cost well above the others, and wage cost at the newest opening was heavier than budget for the whole month. Both problems are now three to six weeks old. The general manager at the city site has already had another month of the same ordering habits, and nobody can remember what happened in week two that pushed wages up.
Everyone agrees the numbers are useful. Nobody can act on them, because by the time they exist the chance to act has passed. So the monthly review becomes a conversation about history, and the real management happens on instinct and on whatever the area managers pick up on their visits.
Why the site P&L is always a month-end job
The data for a site P&L is not missing. It is spread across systems that were never set up to feed one report, and the joining happens once a month in a spreadsheet.
- Sales sit in the EPOS, sometimes more than one EPOS if the group has grown by acquisition.
- Wages come from the rota or payroll system, often only as a monthly payroll figure rather than weekly hours by site.
- Food and drink cost depends on a stock count, and counts are taken on different days at different sites.
- Supplier invoices arrive in a central inbox and are coded to sites in batches, some late, some to the wrong site.
- Delivery platform income and commission come on separate statements that finance reconciles later.
So the finance team waits for the last invoices and the last stock counts before anything is published, because a half-finished P&L invites arguments. That is a reasonable instinct. It just means the report is always late by design.
What a late P&L really costs the group
A cost problem found in week one costs a week of margin. The same problem found on the 17th of the following month has usually cost six weeks, and the habit behind it has had time to set. Menu engineering decisions, rota changes and supplier conversations all wait on a document that is already old.
It also shapes how site managers see head office. When the only feedback on their numbers arrives weeks later, it reads as a telling-off about something they can no longer change. Good GMs want to see their numbers weekly, and many keep their own spreadsheets because the group's report is too slow to help them.
Finance pays too. Someone spends days every month copying figures from exports into the P&L template, fixing site coding and answering the same questions about why a number moved.
How we build a weekly site P&L
We do not replace your accounts package. We build a reporting layer that reads from the systems you already use and produces a site-by-site view on a weekly rhythm, with the month-end accounts still the final word.
- Sales by site, by day and by revenue stream are read from each EPOS through its API or a scheduled export, including delivery platform sales where they come through the till.
- Labour cost by site is built from the rota or time and attendance system, using actual hours worked and the pay rates you provide, not the monthly payroll total.
- Food and drink cost is taken from stock counts where one exists for the week, and from purchases against a theoretical cost where it does not, with the difference clearly labelled.
- Supplier invoices processed in Xero, Sage or your purchasing system are read with their site coding, so a cost lands at the right site in the week it arrives.
- Fixed site costs such as rent, rates and service charges are spread weekly from a simple schedule finance maintains.
- Each line is flagged as actual or estimated, and the view updates when the real figure arrives, so nobody mistakes a provisional number for a final one.
| P&L line | Where it comes from | How firm it is mid-month |
|---|---|---|
| Sales | EPOS by site | Firm, updated daily |
| Wages | Rota or clocking system | Firm for hours, rates as set |
| Food and drink cost | Stock counts or purchases | Estimated until the count |
| Overheads | Finance schedule | Allocated, trued up at month end |
When the month closes, finance still posts the journals and adjustments in the accounts. The weekly view is reconciled back to them so the two never tell different stories for long.
What changes on a Monday morning
On Monday the ops director and each area manager open the same view: every site's sales, wage percentage and estimated GP for last week, against budget and against the other sites. The general manager sees their own site in the same shape. A wage problem at the new opening shows up after the first week, not the fifth.
The monthly pack still happens, but it becomes confirmation rather than news. Finance spends less time building the report and more time explaining the things it shows.
Signs your group needs this
- Site P&Ls land more than two weeks after month end.
- GMs keep their own spreadsheets because the group's report is too slow.
- Invoices are regularly coded to the wrong site and corrected later.
- Nobody can say with confidence how a site traded last week after costs.
- Month-end reporting takes one person several days of copying and pasting.