An investor asks “can you send over your forecast” and a small wave of panic hits. Not because you don’t have numbers , you probably have three different spreadsheets with three different answers , but because you know the real question underneath it. They’re not asking what you think will happen. They’re asking whether you understand your own business well enough to model it.
That distinction changes everything about how you should respond.
The Direct Answer
A financial forecast for investors is a model , usually 3 to 5 years, monthly for year one and quarterly after , that shows revenue, costs, cash burn, and runway, built from clearly stated assumptions rather than a hopeful line trending up and to the right. Investors don’t expect the numbers to be correct. They expect the logic to be defensible. If you can walk someone through why each assumption exists and what would have to be true for it to hold, you’ve passed the test, regardless of whether year three revenue lands at $8M or $11M.
Why Investors Ask for This in the First Place
It helps to sit on the other side of the table for a second. A forecast tells an investor almost nothing about the future , everyone knows early-stage projections are wrong within two quarters. What it tells them is:
- How you think. Do you understand your unit economics, or are you guessing?
- What you believe drives growth. Is it paid acquisition, referrals, sales cycles, network effects?
- Where the money actually goes. Payroll, infrastructure, marketing spend, and whether the ratios make sense for your stage.
- How much runway your ask buys you, and whether that number is realistic or fantasy.
A founder who says “we’ll hit $2M ARR next year” gets a polite nod. A founder who says “we’ll hit $2M ARR if we convert 3% of trial signups at our current CAC of roughly $180, and here’s why we think that conversion rate holds as we scale” gets a follow-up meeting. The forecast is a proxy for founder judgment, not a prophecy.
Bottom-Up vs. Top-Down: Pick the Right Starting Point
Most rejected forecasts fail for one structural reason: they start top-down. “The market is $40B, if we capture just 1%…” Investors have heard this sentence more times than they can count, and it signals the opposite of rigor.
Top-down forecasting starts from total market size and works backward to a market-share assumption. It’s fast, but it’s almost always wrong, because “1% of a huge market” isn’t a strategy , it’s a rounding error dressed up as ambition.
Bottom-up forecasting starts from the actual mechanics of your business: how many customers you can realistically reach, at what conversion rate, at what price, with what retention. It’s slower to build and less flattering, but it’s the version experienced investors actually want to see. If you only have time to build one model well, build this one.
A simple bottom-up SaaS structure might look like:
- Leads or trial signups per month (from your actual channel data, not a guess)
- Conversion rate to paying customer
- Average contract value
- Monthly churn rate
- Resulting MRR, layered month over month with growth and churn netted out
Marketplace, hardware, and services businesses each have their own version of this chain , but the principle holds: start from a unit, and multiply up.
The Assumptions Investors Will Actually Probe
Nobody scrutinizes your revenue formula in a spreadsheet. They scrutinize the three or four assumptions feeding it. Be ready to defend:
- Customer acquisition cost and how it changes as you scale. CAC almost always rises with volume as you exhaust your cheapest channels first. A flat CAC line across 36 months is an immediate red flag.
- Churn or retention curves. Early data from a handful of customers is noisy. Show that you know it’s noisy, and show a range rather than a single confident number.
- Sales cycle length, especially for anything enterprise or B2B. Founders routinely underestimate this by half.
- Hiring plan and its lag on productivity. New hires , especially in sales , take months to become fully productive. Modeling them as instantly effective inflates every downstream number.
- Gross margin, particularly if there’s any hardware, hosting-heavy infrastructure, or services component riding along with the “software” story.
You don’t need certainty on any of these. You need a stated range, a reason for the range, and evidence of where the number is likely to land as you get more data.
Structuring the Forecast Itself
A forecast that survives scrutiny usually has three linked layers, not one tab of formulas:
Revenue model. Built bottom-up as above, broken out by product line or customer segment if you have more than one. Show the assumptions on a separate, clearly labeled tab , don’t bury them in cell formulas where no one can see your logic without clicking through.
Operating expense model. Headcount is almost always the dominant cost at early stage. Build a hiring plan by role and month, tie salary and burdened cost (benefits, payroll tax, roughly 1.25–1.4x base salary as a rule of thumb) to each hire, then add non-headcount costs: tools, hosting, office, marketing spend as a percentage of revenue or a fixed monthly number.
Cash and runway. This is the tab investors flip to first. Starting cash, monthly burn, months of runway remaining, and , critically , what the requested investment buys in terms of additional runway and what milestones you’ll hit before needing to raise again. If your ask doesn’t map cleanly to a runway extension and a specific set of milestones, that mismatch will get noticed immediately.
Common Mistakes That Undermine an Otherwise Good Forecast
The hockey stick with no mechanism. Growth that suddenly accelerates in month 18 with no explanation for why , no new channel, no product launch, no pricing change , reads as a plug to hit a round number, because it usually is one.
Ignoring your own historical data. If you have six months of actual numbers, your forecast needs to reconcile with them. A model that assumes 15% month-over-month growth going forward when your trailing three months averaged 4% needs a very good explanation for the gap.
One scenario instead of three. A single-point forecast implies false precision. A base case, an upside case, and a downside case , even a simple version with two or three flexed assumptions , shows you understand your own model’s sensitivity rather than treating it as gospel.
Confusing bookings, revenue, and cash. These are three different numbers and conflating them is one of the fastest ways to lose credibility with anyone who’s read a cap table before.
Presenting the forecast without the assumptions page. The spreadsheet with formulas is not the deliverable. The one-page summary of “here are our five key assumptions and why we believe them” is what actually gets read closely.
What Tools You Use Matters Less Than the Logic
Google Sheets or Excel, built and fully owned by the founding team, is genuinely fine and often preferred , investors want to see that you built it and understand every cell, not that a fractional CFO assembled something impressive-looking you can’t defend live. Purpose-built forecasting tools (several exist for early-stage SaaS financial modeling) can speed up formatting and scenario toggling once you’re past seed stage and juggling more complexity, but they don’t replace the underlying thinking. Check current features and pricing directly before choosing one, since this space changes quickly. No tool substitutes for you being able to explain, unprompted, why revenue grows the way it does in month 14.
Frequently Asked Questions
Three years is standard for seed and Series A. Year one monthly, years two and three quarterly or annually. Anything beyond three years for an early-stage company is closer to fiction than modeling, and most investors know it.
Say so directly and explain why, then show the updated model. Investors respect a founder who owns a miss and adjusts more than one who quietly reshuffles assumptions to make the new chart look consistent with the old one.
Yes, in broad terms. A downside case with a clear-eyed view of when you’d need to cut costs, change strategy, or wind down shows judgment. Avoiding the topic entirely reads as avoidance, not confidence.
No. They expect it to be well-reasoned. Nearly every early-stage forecast misses its numbers within a year or two. What investors are actually evaluating is whether you’ll notice the miss quickly, understand why it happened, and adjust , which is a proxy for how you’ll run the company after the check clears.
If a single tab needs a guided tour to understand what’s happening, it’s over-built. Aim for a model any team member could open and reconstruct your logic from within a few minutes, backed by a one-page written summary of assumptions for anyone who won’t open the spreadsheet at all.
Treating the forecast as a sales document instead of a thinking document. The version that impresses investors most is rarely the version with the biggest numbers , it’s the version that makes clear you’d build the same model, with the same rigor, whether or not anyone outside the company ever saw it.
Bringing It Together
A forecast isn’t a prediction of the future , it’s evidence of how carefully you think about your own business. Investors have seen thousands of these decks and models, and they can tell within minutes whether the numbers were built from real unit economics or reverse-engineered from a target raise. Build bottom-up, state your assumptions in plain language, show more than one scenario, and be ready to explain every number out loud without the spreadsheet in front of you.
Get that right, and the actual figures in the forecast matter far less than most founders assume they do.
