Startup Financial Projections: A Plain-English Guide

Startup financial projections are your best estimate of how much money comes in, how much goes out, and how much cash you'll have left, month by month, usually for the next three years. They're not predictions. They're a model of your assumptions, so you can see what has to be true for the business to work, and what happens if it isn't.

That shift in thinking matters. You aren't trying to guess the future correctly. You're trying to make your guesses visible, so you (and anyone you show them to) can argue with them.

What financial projections include

A complete set has three parts that connect to each other:

• Profit and loss (income statement): revenue, costs and whether each month made or lost money.
• Cash flow: when money actually arrives and leaves. This is the one that tells you whether you survive.
• Balance sheet: what the business owns and owes at each month end. It must balance every month, which is a handy check that the model isn't broken.

Most early-stage plans show year one month by month, then years two and three by quarter or year. If you're raising money, add a short set of KPIs: monthly recurring revenue, burn, runway, gross margin, and customer acquisition cost.

Start with drivers, not a revenue number

The fastest way to build useless projections is to start with the answer: "We'll make 1 million in year three," then fill in whatever makes it true. Anyone who reads plans for a living will spot it.

Build from the bottom up instead. Revenue is the end of a chain:

1. How many people see your product each month, per channel
2. How many of them sign up
3. How many sign-ups start paying, and after how long
4. How much each customer pays
5. How many customers cancel each month

Each link is something you can test, measure and improve. When a projection is built this way, a question like "what if conversion is half what you think?" has an instant answer.

The seven assumptions that matter most

For most startups, these few numbers drive nearly everything:

Write a one-line note next to each: where did this number come from? "Guess" is an allowed answer. Hiding that it's a guess isn't.

Build it month by month

Here's a simple chain for one month, using the examples above:

• 800 visitors × 3% = 24 sign-ups
• 24 sign-ups × 5% = 1.2 new paying customers
• Customers at month end = last month's customers × (1 − 5% churn) + new customers
• Revenue = customers × 19

Run that forward for 36 months and add growth to the visitor numbers. You'll notice something humbling: small early numbers take a long time to compound. That's normal. It's also why a plan that shows 10,000 customers in month six raises eyebrows.

Costs work the same way. Split them into fixed costs (the same every month) and variable costs (they grow with customers). Hosting, AI usage and payment fees are usually variable. Team costs are fixed until you hire.

Revenue is not cash

This is where many first-time models go wrong. Revenue is when you earn money. Cash is when it lands in your account.

• An annual plan brings twelve months of cash upfront, but counts as revenue month by month. The difference sits on the balance sheet as deferred revenue.
• A lifetime deal brings cash once, often a month or two after the sale, and then costs you money to serve for years.
• Marketplaces and payment providers can pay out weeks after the sale.

Your cash flow statement is the one that decides whether you make payroll. When in doubt, plan with cash.

How investors read your projections

Experienced investors rarely believe the totals. That's fine, they don't expect to. What they're checking is whether you understand your business:

• Are the assumptions sensible? A 40% sign-up rate or zero churn tells them you haven't looked closely.
• When do you break even, and how low does cash go before that? This is your funding need.
• Does each customer make money? Compare lifetime value with acquisition cost.
• What happens if growth is half your plan? Having a slower scenario ready builds a lot of trust.

A good line to use in any pitch: "These are projections from these assumptions, and here's how they change if we're wrong."

Sanity checks before you share them

Run these quickly before anyone else sees your model:

• Does monthly growth ever exceed what your channels could realistically produce?
• Is your year-three revenue a believable share of the market you can actually reach?
• Is gross margin in a normal range for your kind of business?
• Is the team big enough to handle the customers in the plan, including support?
• Does the balance sheet balance every month?

If a check fails, fix the assumption, not the total.

Keep them alive

Projections are most useful after launch. Each month, enter what actually happened: revenue, costs and customers. The gaps tell you which assumption was wrong. When actual numbers drift far from the plan, re-forecast from reality instead of defending the old plan.

A faster way to build them

You can build all of this in a spreadsheet, and it's a great way to learn. If you'd rather skip the formulas, startzero.money builds the full set (profit and loss, cash flow, balance sheet and KPIs) from six plain questions, keeps the three statements connected, and shows how every number is calculated. Try the free runway calculator or break-even calculator first if you only need one answer.