How to Create a Cash Flow Forecast for Startups That Wins Investors

How to create a cash flow forecast for startups (without lying to yourself)

The first time I built a cash flow forecast, I gave myself 14 months of runway. I actually had six. I found out on a Tuesday, staring at a bank balance that didn't match my beautiful spreadsheet, and I spent the rest of that week cancelling things I'd signed up for three months earlier.

That's the whole problem with how to create a cash flow forecast for startups: it's not a math exercise. It's an honesty exercise. CB Insights analyzed 431 startup failures and found roughly 38% died from running out of cash — more than any other single cause. Not bad ideas. Not competitors. Running out of money while the spreadsheet still said everything was fine.

So here's what I actually do now, after building these things for my own companies and for a handful of founders who asked me to look at theirs.

Why this matters more than your revenue projection

A P&L can look great while your bank account is empty. This is the part that trips up almost every first-time founder, and I include myself in that group.

Revenue is recognized when you earn it. Cash arrives when the client decides to pay. Gap between the two? For a small B2B startup, that's often 45 to 75 days — and that gap is where startups die quietly, with a healthy-looking pipeline and no money to make payroll.

Cash is not revenue, and this distinction will save you

Say you invoice a client €20,000 on March 1st with net-60 terms. Your accounting software might show that revenue in March. Your bank shows it in May. If you planned February's hiring around that March number, you're paying salaries with money that doesn't exist yet.

What I track, and what I'd tell any founder to track, is the cash conversion cycle: how many days pass between spending money and getting it back. The longer it is, the more working capital you need to survive growth. Growing fast with a slow collection cycle is a fantastic way to go bankrupt while looking successful.

The statistic nobody wants to hear

CB Insights' post-mortem of failed startups puts "ran out of cash / failed to raise new capital" at the top of the list, cited in 38% of the failures they studied. Startup Genome's older work on 3,200+ companies put the number even higher. Different methodologies, same conclusion: it's the default ending.

Here's the thing though. Running out of cash is almost never a surprise when it happens. It was visible in the numbers three to five months earlier. The founder just didn't have a forecast granular enough to see it.

Building your forecast, step by step

There are plenty of 4-step guides out there that tell you to "estimate inflows, estimate outflows, consolidate." Technically correct. Practically useless, because they skip the part where you have to decide what you actually believe.

Start with assumptions, not numbers

Before you open Excel or a template, write these on a page:

• How long until a new customer pays you? Ask your actual customers. Don't guess from an invoice template.
• What's your committed monthly spend — the stuff you cannot cancel next week without a penalty?
• When does your next funding round land, and how confident are you in that date? (Spoiler: cut it by 30-40%.)
• Which expenses scale with revenue, and which ones are fixed no matter what?
• What was your actual burn over the last three months? Not the projected burn — the real one, pulled from bank statements.

I cannot overstate this last one. My first forecast was optimistic by roughly 25% on the expense side, mostly because I forgot annual software renewals, one-off legal invoices, and the fact that I always spend more on contractors than planned.

Build the grid, weekly not monthly

Monthly columns are standard. They're also not good enough. A cash forecast for startups should have weekly columns for the first 13 weeks, then monthly beyond that. Thirteen weeks is enough resolution to catch a problem before it's fatal.

Your sheet needs, at minimum:

1. Opening cash balance for the period
2. Cash in — customer payments by expected date, not invoice date
3. Cash out — payroll, rent, tools, contractors, VAT/tax, loan repayments
4. Net movement = cash in minus cash out
5. Closing balance = opening balance plus net movement
6. Runway in weeks = closing balance ÷ average weekly burn

That last line is why you do this at all. Everything else is bookkeeping. Runway is the number you check every Monday morning.

Wells Fargo's small business guidance points at the same four-step structure most banks use: identify assumptions, estimate cash receipts, estimate cash disbursements, then consolidate. The difference is whether you do it weekly and honestly, or monthly and wishfully.

If you have no revenue yet, your forecast is different

Every template you'll download assumes you have customers. Most pre-seed startups don't. This is the gap nobody fills properly, so let me be specific.

For a pre-revenue company, your forecast isn't about inflows and outflows. It's about runway and fundraise timing. Two variables.

Burn rate and the round that has to land

Your monthly burn is your total monthly cash outflow, minus any small revenue you do have. Take the last three months, average them, and add 10% because costs creep. That's your baseline burn.

Now the fundraise. If your forecast assumes a close on, say, June 1st, model it as August 1st. Legal docs, diligence, international wires — all of it takes longer than founders think. I've seen nine-month rounds that everyone was sure would close in three.

Sketch it like this:

• Cash on hand today: €X
• Monthly burn: €Y (including your own salary at a real number, not the founder-discount version)
• Months of runway without new money: X ÷ Y
• Fundraise start month: 6 months before your runway runs out — the minimum time to raise a seed round in most markets
• Buffer: two months on top, because things slip

If X ÷ Y gives you 14 months, and raising takes 6, you have 8 months of real work time. That's your deadline, whether you like it or not.

Scenarios, templates, and the tools I actually use

One forecast is a guess. Three forecasts — best, base, worst — turn a guess into a decision framework.

Best, base, worst — and which one to trust

Scenario	What it assumes	Planning use
Best case	Customers pay on time, one big deal lands, fundraising closes early	Good for investors. Do not plan around it.
Base case	Realistic payment delays, no surprise wins, round closes 30% late	Your working plan
Worst case	Two months of delayed payments, one key hire doesn't onboard, round slips a quarter	Your actual decision-making model. Yes, really.

I'll die on this hill: build your hiring plan and your spending plan around the worst case. If your base case comes true, you have unspent money — which is a great problem to have. If your best case was your plan and it doesn't happen, you're cutting salaries at 2am on a Saturday. Choose which problem you'd rather have.

Templates and tools — what's worth using

For how to create a cash flow forecast for startups in Excel, you don't need anything fancy. Excel, Google Sheets, and Apple Numbers all handle it. I've used all three. Google Sheets wins purely because I can share a live link with a co-founder and a board member and never email a stale file again.

Free templates exist from Wells Fargo, the British Business Bank, SCORE, and a dozen accounting blogs. Grab one, delete half of it, and rebuild the columns and rows to match your actual business. Templates are for structure, not for guessing your numbers.

Paid tools — Forecastr, Runway, Finmark, and similar — are worth it once you have revenue above roughly €1M ARR and multiple departments writing expenses. Below that, you're paying for sophistication you won't use. I cancelled two of these subscriptions after three months each and never missed them.

One thing no template will do for you: check the actual bank account. Every Monday, I open the bank statement and reconcile. Five minutes. Catches errors that would otherwise compound for a month.

Making the forecast honest

Forecasts fail for behavioral reasons more than technical ones. Here are the traps I've fallen into personally.

• Optimism on collection timing. Clients say "we pay in 30 days." Reality is 50+.
• Vanishing expenses. Software you forgot you subscribed to. Annual renewals that surprise you. My worst one was a €4,200 compliance filing I had completely forgotten was annual.
• Founder salary at zero. Paying yourself nothing isn't a strategy — it's a forecast error that shows up as personal debt later.
• Fundraise timing fantasy. If you've never closed a round, add 50% to whatever timeline your investor friend quotes.

The fix isn't discipline. It's a system: update the forecast every Monday, compare it against the previous week's projection, and write down why the numbers moved. After eight weeks, your model becomes meaningfully accurate — not because the future got clearer, but because you now know your own error patterns.

I still miss. My forecasts run about 12-15% off from actuals in a typical month, which is dramatically better than the 40% miss I was getting in year one. That improvement didn't come from a better template. It came from tracking why I was wrong.

So the honest question isn't whether your first forecast is accurate. It's whether you'll update it next Monday when it isn't. That's the difference between the 38% who run out of cash, and the ones who see it coming and do something about it.