Audit
- Money and processes across all sources
- Report, tables and 16 charts
We’re moving the chain to a system that collects sales, supplier invoices, bank payments and delivery reports into one database every night. In the morning the owner gets the numbers in a dashboard and in Telegram — no spreadsheets.
stage 1 in progress| Revenue | 9,812,000 | 100% |
| Food cost42 invoices — open › | −3,061,000 | 31.2% |
| Payroll | −2,158,000 | 22.0% |
| Rent | −1,079,000 | 11.0% |
| Delivery commissions | −834,000 | 8.5% |
| Ads | −294,000 | 3.0% |
| Other | −1,075,000 | 11.0% |
| Net profit | 1,311,000 | 13.4% |
Restaurant N · Thu, 15.10
Turnover: 412,300 (+6% vs a usual Thursday)
Center — 248,100: dine-in 171,400 (96 checks), Wolt 76,700 (41 orders)
New Belgrade — 164,200: dine-in 118,900 (71 checks), Wolt 45,300 (24 orders)
Month to date: 5.9M (+4% vs 1–15 Sep)
Attention
• New Belgrade: Wolt orders 24 — −27% vs usual
• Large invoice: fish supplier — 186,000 (usually ~120,000)
• 2 invoices unpaid after due date
Data: POS, banks, e-invoices — as of 15.10.
09:00Illustrative numbers. The client’s data stays private.
of bank payments matched to supplier invoices
sales per dinar of delivery ads — one location vs the other
records from 5 systems in one model
when the second location paid back — seen in the data
Every night, on its own
The owner used to put the numbers together by hand. The system makes them arrive on their own.
From the first audit to a system that runs every day.
Each stage was delivered and accepted before the next one started.
Found a supplier invoice paid twice, and which delivery ads pay off 10× — and which don’t at all.
Every data source mapped: where it lives and what can be automated.
Nightly data, P&L, dashboard and the morning summary.
Two new dashboard sections: guests and shifts.
The first call is 30 minutes and free.