SME owners run their business on three senses: the meeting calendar, the bank balance, and "I have a feeling it'll be tight". The first two are reliable. The third one is not. And yet the third one decides whether, halfway through the month, you'll be knocking on the bank's door for a credit line.
A cash flow forecast — even a simple one, 13 weeks ahead, refreshed once a week — is to a small or mid-sized business what a speedometer is to a car. It doesn't tell you where you're going. It tells you how far you can still go before you have to stop. You'd be surprised how few companies under 50 people actually have one.
The reason is simple. All the data exists, it's just scattered across the ERP, the bank and the accountant's head. Nobody wants to spend three hours every Monday morning gluing it together in Excel — and we don't blame them.
13 weeks is enough
Forget for a moment "strategic financial planning" three years ahead. That's a different animal. The operational forecast we're talking about has one job: to show you, 13 weeks out, in which week you'll run out of cash if nothing changes.
Why 13 weeks? Because that's the window where you can still do something about it — pull an invoice forward with a customer, push an order out with a supplier, renegotiate a term, delay buying that new van. Any shorter, you won't have time to react. Any longer, the data is too uncertain to plan on.
The structure we recommend to clients fits on a single A4 page:
- Expected inflows: sales invoices with a payment date in the given week.
- Confirmed inflows: the same, but flagged as "customer promised to pay".
- Fixed outflows: payroll (including social security and PIT-4), rent, leases, subscriptions.
- Variable outflows: purchase invoices due in the given week.
- Taxes: VAT by the 25th, corporate/personal income tax by the 20th, social security by the 15th or 20th.
- Weekly balance and running balance.
Nothing more. It has to fit on one screen so the owner can glance at it over Monday morning coffee.
Where the data comes from
If the company runs on Comarch Optima, enova365, Subiekt GT or Microsoft Dynamics AX, everything you need for inflows and outflows is already in there. Unpaid sales invoices, unpaid purchase invoices, payroll schedule, tax calendar — these are standard reports, they're just spread across different screens.
In practice it comes down to four sources:
- ERP — unpaid sales invoices (receivables), unpaid purchase invoices (payables).
- HR/payroll — net payroll schedule plus social security and PIT-4 contributions (from an HR module or a separate system).
- Bank — current balance and recurring transfers (leases, subscriptions, rent).
- Tax calendar — VAT, corporate/personal income tax advance, social security. Four dates in a month, always known in advance.
Gluing these four sources together by hand takes an average accountant 2 to 4 hours a week. Automating them — hooking up the ERP via API or at least a scheduled export, plus the bank statement via PSD2 — brings it down to a 20-30 minute review and sign-off.
Three traps everyone falls into
Trap one — "paper" receivables. The company issued an invoice with a 30-day term and dropped it into the forecast as an inflow in week five. Meanwhile that specific customer pays on average in 47 days. The forecast shows +40,000 zł that isn't actually there. The fix — don't take the term off the invoice, take the average payment time per customer. That's three SQL queries in the ERP; you do it once and you get a realistic picture.
Trap two — VAT as "money passing through". VAT is not your cost, it's yours temporarily. But in cash terms, it's a real outflow. Companies that run their forecast net regularly find out on the 25th that they don't have enough to pay VAT. The forecast has to be gross, with a clearly marked VAT-payable line.
Trap three — seasonality seen through a weekly lens. If your business does 40% of its turnover in November–December, a forecast run in February will show that summer looks tight. That's when people start making rushed decisions. A cash flow forecast doesn't replace the annual budget — it's operational, not strategic. Better to say that out loud so nobody panics quarterly for no reason.
A real example
At one of our clients — a 22-person electrical wholesaler running Subiekt GT and a separate Excel sheet — the cash flow forecast was maintained by the head accountant. Every Monday, updating it took her about 3.5 hours: export from Subiekt, manual paste of the ING Business bank statement, checking payroll, adding the VAT/social security calendar. The forecast reached 4 weeks ahead — beyond that "too much was moving".
After we connected Subiekt through its SFERA API to the report and pulled the bank statement over PSD2, the same job now takes 25 minutes on Monday morning. And the forecast reaches 13 weeks — because the data refreshes overnight and the accountant only reviews the "unusual" items.
A side effect that surprised the board: in the first quarter, the forecast twice flagged a risk of a negative balance in a specific week. Both times it was resolved by pulling an invoice forward with a large customer (a phone call to their AP, 0.5% discount for early payment). Without the forecast they'd have found out when the social security transfer bounced.
Build it yourself or outsource it
If someone in the company actually knows Excel and you have API access to the ERP, you can do it yourselves in 3-5 days of work. If you don't have that person, or the ERP's API isn't documented (which happens a lot with older versions of Optima or Subiekt), it's faster and cheaper to buy it as a one-off project from outside. We're talking about a 15,000-40,000 zł engagement, not a year-long implementation.
Don't buy dedicated "liquidity management" software for 800 zł a month — you need one good report, not another system to log into.
A cash flow forecast doesn't give you more money. It tells you about the problem six weeks before it happens — and that's usually enough to solve it.