Your startup can be profitable on paper and still run out of money on a Tuesday afternoon. I've watched it happen twice — once to a company I consulted for, once almost to my own. Both times the P&L looked fine. Both times the bank account told a different story.
That gap between profit and cash is why a cash flow forecast for startups isn't a nice-to-have. It's the thing that tells you whether you can make payroll in eight weeks or whether you need to start that funding conversation now, not when the account hits zero.
Here's how to build one that actually holds up.
Key Takeaways
- Forecast cash, not profit. They are not the same number, and confusing them kills startups.
- Use a 13-week rolling forecast for near-term survival and a 12-month view for planning.
- Build from real drivers (headcount, contracts, payment terms), not from a flat growth assumption.
- Always run three scenarios: base, pessimistic, optimistic. The pessimistic one is the one you plan against.
- Update weekly. A forecast you touch once a quarter is decoration.
- Track your runway and your minimum cash threshold religiously.
Why a cash flow forecast matters more than your P&L
Your P&L records revenue when you earn it. Your bank records cash when it lands. Those two events can be 90 days apart, and in that gap, startups die.
Consider a fairly normal deal: you sign a $60,000 annual contract in January, invoice it, and the client pays net-60. On your income statement, you've booked revenue. In your bank account, nothing shows up until March. Meanwhile, you've paid your engineers, your cloud bill, and your office rent for two months. You were "profitable." You were also broke.
This is why I stopped trusting anyone's profit figure when they asked me for advice on runway. The number that matters is simple: money in, minus money out, over time. Everything else is commentary.
The two horizons you actually need
You don't need one forecast. You need two, serving different jobs.
- 13-week rolling forecast — weekly granularity, for survival. This catches the short-term crunches before they become emergencies.
- 12-month forecast — monthly granularity, for planning hiring, fundraising, and product bets.
Some founders also keep an 18-month view when they're heading into a raise, since investors want to see you can survive the round plus a buffer. But start with those two.
How to build your first forecast, step by step
You don't need a finance degree. You need a spreadsheet, your last three months of bank statements, and brutal honesty about your collection timelines.
Start with cash in
Cash in comes from three places: customer payments, financing (loans, grants, equity), and anything else (tax refunds, asset sales). For most startups, customer payments are the hard part.
The mistake I see constantly is forecasting revenue instead of collections. If a client pays net-30, that money doesn't help you on day 15. Build a collections schedule:
- List every invoice with its issue date and payment terms.
- Apply your actual average collection period — not the terms, the reality. If your customers say net-30 but pay in 45 days on average, use 45.
- Spread the collections across the weeks they'll actually land.
Track your DSO (days sales outstanding) religiously. One client paying late can move it from 30 to 55 days, and that shift alone can gut a quarter.
Then cash out
Cash out is easier to predict because you control most of it. Group it into categories:
- Payroll and contractors (your biggest line, usually 60-70% of spend)
- Software and infrastructure
- Rent and utilities
- Marketing spend
- Professional services (legal, accounting)
- Taxes — VAT/GST, payroll taxes, corporate tax. Do not forget these. I've seen founders wipe out a month's buffer by forgetting quarterly VAT.
- Loan repayments and interest
For each line, decide whether it's fixed (same every month) or variable (scales with activity). Fixed lines are easy. Variable lines need a driver — headcount, traffic, order volume.
Net it out and find your runway
Subtract cash out from cash in for each period. Add the result to your opening balance. That running total is your projected cash position.
The lowest point on that curve, before it recovers, is your cash trough. That's the number that determines whether you're safe or whether you have a problem. And runway — how many months until you hit zero at your current burn — is just that trough expressed in months.
One number I'd tattoo on every founder's forearm: keep at least three months of operating expenses as a minimum cash threshold. Below that, you're not running a business, you're gambling.
Which tool should you use?
You can go a long way with a spreadsheet. But at some point, the manual updates eat your time and introduce errors. Here's the honest comparison.
| Tool type | Best for | Weakness | Cost range |
|---|---|---|---|
| Spreadsheet (Excel/Sheets) | Pre-revenue to early revenue; total control | Manual, breaks with scale, no bank sync | Free |
| Integrated finance platforms | Startups with recurring revenue and multiple accounts | Setup time, subscription cost, occasional sync errors | Mid-tier monthly |
| Accounting add-ons | Teams already living in their accounting software | Limited scenario modeling | Low monthly |
| Full FP&A suites | Series A+ with a finance hire | Overkill and expensive for pre-seed | High monthly |
My honest take: start with a spreadsheet. Move to a platform when updating your forecast takes more than an hour a week, or when you have more than three revenue streams. Not before. I've seen pre-seed teams burn budget on finance software they never opened.
Scenarios and stress testing
A single forecast is a guess wearing a suit. Three forecasts give you range.
The three scenarios
Build these on top of each other:
- Base — your realistic expectation. This is the one you hope for.
- Pessimistic — slower sales, longer collections, one big client churning. Plan against this one.
- Optimistic — everything lands. Use it to understand upside, not to make decisions.
The pessimistic scenario is your survival plan. If your company can't survive the pessimistic case, you need to cut costs or raise now. Not in three months. Now.
Watch your burn multiple
One metric worth tracking alongside cash: your burn multiple — net cash burned divided by net new recurring revenue added. If you're burning $3 to add $1 of new recurring revenue, that's a warning sign, regardless of how healthy the bank account looks today.
The threshold varies by stage. But if that ratio climbs quarter over quarter, your growth is getting more expensive, and your forecast should reflect it.
Common mistakes that wreck a startup forecast
I've made most of these. Learn from my bruises.
Confusing profit with cash
Covered above, but it deserves repeating because it's the single most common error. Revenue recognized is not cash received. Never forecast one as if it were the other.
Forgetting taxes and payment timing
VAT, payroll taxes, and quarterly estimated taxes hit at specific times. Miss them in your forecast and you'll get an unpleasant surprise. I once watched a founder discover a five-figure VAT bill two days before it was due. His forecast had it in the wrong month.
Double-counting recurring revenue
If you forecast a subscription contract's full annual value in month one, you've overstated cash by eleven months. Subscription revenue arrives in slices. Model it that way.
Assuming best-case collections
Your best customer pays on time. Your worst one doesn't. Average them, and skew toward the pessimistic. If you've never tracked DSO, do it before you build the forecast. You might be shocked.
How often should you update your cash flow forecast?
Weekly for the 13-week view. Monthly for the 12-month view. This is the part most guides skip, and it's the part that matters most.
Why weekly? Because cash events move fast. A payroll date shifts, a client asks for an extension, a big invoice clears early. If you only review monthly, you're always reacting. Weekly review means you spot the crunch with time to act.
The update itself shouldn't take long. Fifteen to thirty minutes once the structure exists. You're plugging in new actuals and pushing the horizon forward one week. That rolling motion is what keeps the forecast alive.
And here's the discipline that separates founders who sleep from founders who don't: every week, compare your forecast to what actually happened. Where you were wrong, ask why. That feedback loop sharpens the forecast faster than any tool.
A final thought on forecasts you'll actually use
A cash flow forecast won't fix a broken business model. It won't make a slow-paying client pay faster. What it does is give you something rarer: time. Time to cut costs before you're forced to. Time to start a raise before you're desperate. Time to say no to a deal that would stretch your cash to breaking point.
Two years ago I helped a founder rebuild his forecast from scratch. He'd been running on gut feeling and a bank balance he checked every morning with a knot in his stomach. Six weeks after we built a proper 13-week rolling model, he spotted a trough coming in eleven weeks. He had time. He negotiated a payment plan with one vendor, pushed a hire back a quarter, and pulled a deposit forward. Crisis absorbed.
The forecast didn't save his company. It just showed him where the cliff was, early enough to walk around it.
That's all it ever needs to do.
