Most pre-revenue forecasts are fiction dressed up as spreadsheet discipline. The revenue curve looks heroic, the costs look tidy, and nobody in the room believes either one. Investors aren't checking whether you'll be right about month nine. They're checking whether you understand how cash actually moves through your business, and whether the plan bends sensibly when reality pushes back. A 12-month cash flow forecast is the cheapest way to prove that.
A cash flow forecast tracks when money enters and leaves your bank account. An income statement tracks when revenue is earned and costs are incurred. Before you have revenue, that distinction matters less on the top line and much more on costs: a $12,000 annual software plan hurts in the month you pay it, not in twelve gentle slices.
Build one file with two tabs. The first holds your assumptions — headcount, salaries, rent, subscription costs, expected close dates, payment terms. The second holds the forecast, where every number is a formula pointing back at those assumptions. Rows should run from opening cash through cash in, each cost category, net burn, and closing cash. Columns run across twelve months, and each month's closing balance becomes the next month's opening balance. If the forecast tab is full of hard-coded numbers, you've made a poster, not a model.
One naming discipline helps more than any formatting: call the bottom line net burn, and label runway as months of cash left at the current burn. Those two numbers anchor every investor conversation you'll have.
Fixed costs are the part you can predict with confidence, which is exactly why they deserve the most rigor. Open the last three months of bank and card statements and go through them line by line instead of from memory. Then sort everything into a handful of categories:
Watch for costs that don't arrive monthly. Insurance often bills annually or quarterly, legal invoices land in lumps after a financing, equipment gets bought in bursts. Put them in the month they'll actually hit the bank, and keep a small "known irregulars" line so a renewal you forgot about doesn't detonate month seven.
The most common flaw in a startup forecast is hiring that appears because the founder wants it to, not because the business has earned it. Replace dates with triggers. A trigger is a specific, checkable event that releases the spend.
Examples that work for a pre-revenue company: hire the first engineer once the prototype handles a defined workload; bring on a designer once three design partners have committed to a pilot; add a salesperson once the founder has closed two deals without help. Attach each hire to its trigger in the assumptions tab, then let the cost switch on only when the trigger condition is marked true.
If a hire has no trigger, it has no place in the forecast.
Budget the fully loaded cost, not the salary. Employer taxes, benefits, equipment, and recruiting fees typically add meaningfully to base pay, and the exact percentage varies by country and contract type — check with an accountant in your jurisdiction rather than guessing. Model a four-to-six week gap between opening a role and the first paycheck, and give yourself a one-month buffer on top. Then sequence hires instead of stacking them: two people who start in the same month cost twice as much in that month and every month after.
Your forecast and your raise are the same story told twice. A 12-month model should show the month you expect cash to land, the runway it buys, and which milestones it funds. Three habits keep that honest.
Match the use of proceeds to the forecast. If you tell an investor the round funds six hires, those six hires should be visible on specific months with specific triggers. Mismatches between the pitch and the model are the fastest way to lose trust in a diligence call.
Three versions of the same model are enough: a base case with realistic timing, a slow case where costs stay flat and the raise slips, and a down case where you cut discretionary spend and hire only against funded triggers. The slow case is usually the one that saves the company, because it tells you the exact month you'd need to act.
The rule that keeps all three credible: no cash appears in the forecast until it's contracted. A pilot that "should" close in March goes in April at the earliest, and only if a signed agreement exists. Pre-revenue projections built on hopeful revenue lines collapse under one question from an investor — "who has committed?" — and take the rest of your numbers down with them.
A forecast that's built once and never touched is a fundraising prop. Make it a rhythm instead. Check actual cash weekly against the forecast, and add a variance column so you can see which lines drift. Reforecast at the end of each month using real numbers for the months behind you, and note what changed and why. Patterns matter more than single misses: three months of hosting costs running 40% over plan is a signal about your architecture or your pricing, not bad luck.
Keep the model short enough to explain in five minutes. One page of assumptions, one page of monthly output, one page comparing the three cases. Save the detailed cost schedules for backup.
When you sit down with an investor, bring the assumptions tab, not just the summary. Say plainly: here's our cash today, here's the burn, here's what a raise buys us, and here's the trigger that tells us to spend more or pull back. Then invite the stress test. Expect questions about the timing of your first hire, the costs that could surprise you, and what happens if the round takes two quarters longer than planned. Having a prepared answer to each is worth more than a perfect model.
Keep the file versioned and dated, and revisit it after every financing and every material change in plan. If the numbers stop matching reality, fix the assumptions rather than the output. This article is general guidance; for tax, payroll, and equity questions specific to your company, work with a qualified accountant or lawyer.
Photo: Alesia Kozik / Pexels
