Startup financial model: how to build one you can defend

Picture a classic first model. Revenue takes off like a ski jump in month 13, nobody can say why, and the team triples without a single line for social contributions. An investor would spot both problems in a couple of minutes.

This guide shows you how to build a startup financial model that survives that first read: the six parts of a 36-month model, why you go monthly, why you need three scenarios, what an employee really costs in Italy, and a small invented SaaS worked out month by month. New to all this? Start from the beginner’s map from zero to your first round.

What a startup financial model is for

At pre-seed and seed, nobody believes your month-30 revenue. Fine. This model isn’t a forecast anyone will hold you to. It shows how you think, and it tells you how much to raise and when to hire.

Day to day, it settles three questions. How much cash gets us to the next milestone? Which assumptions is everything hanging on? And if one of them is wrong, what happens to the cash?

It also produces the numbers investors ask about. MRR, the monthly recurring revenue of a subscription business, comes out of it, and so does your runway, the months of cash you have left. If you’ve never worked those out for a simple budget, read how much money you need to start a startup first.

The six building blocks of a 36-month model

1. Assumptions. Everything you type in lives on one sheet: price, conversion rates, churn (the share of customers who leave each month), salaries, hiring dates, rent. Give inputs their own colour. In Christoph Janz’s well-known SaaS plan template they’re blue, which means anyone can change a cell and watch what moves.
2. Revenue drivers, built bottom-up. Ask where each customer comes from. Maybe a marketing budget divided by your CAC (what it costs to win one customer), maybe a salesperson’s monthly capacity, maybe referrals. Then multiply by price and take away the people who leave. What’s left is MRR.
3. Headcount plan. Every person, their role, start month and gross salary, plus employer costs. In the example below it’s about two thirds of all costs by month 24.
4. Operating costs. The accountant’s invoice, a coworking desk, software subscriptions, travel, insurance, marketing. None of them looks big. Together they’re a real number.
5. Cash flow. Here you record money when it actually moves, not when you send the invoice. A client who pays at 60 days puts the cash in your account two months later. Funding rounds and grants go on this sheet too.
6. Runway. Take the cash left at the end of a month and divide it by that month’s net burn. Every other block feeds this one number, and it answers the question that matters: when does the money run out?

Even Janz didn’t claim his template fitted every business. He called it a starting point, simplifications included. Take the hint and keep yours lean: fewer lines, and logic anyone can follow.

Monthly or annual? And why you need three scenarios

Build all 36 months as months. A yearly total hides the thing you most need to see, the month your cash is at its lowest, and a hire in March instead of October can move that month by several months.

Then make three copies. The base case is what you honestly expect. In the downside, customers cost more to win and leave sooner. In the upside, things go better than planned. Don’t touch the structure between copies, change only the drivers, or you can’t compare them.

Thirty-six months is not arbitrary. A round has to carry you to the next one, and that stretch is long: Carta’s US data put the median time between a seed round and a Series A at 1.9 years at the end of 2025. Three years puts the round, the next milestone and the next raise on one page.

A worked example: an invented SaaS, month by month

Take a deliberately small example. The startup is in Turin and sells booking software to physiotherapy studios. Pre-seed has just closed, with €400,000 in the bank. Its assumptions sheet says:

• Price: €69 a month per studio.
• New customers: marketing spend divided by a CAC of €700, with the budget at €2,000 a month until month 6 and €5,000 after that. Referrals add 3 a month. A salesperson joins in month 13 and ramps up to closing 6 a month.
• Churn: 3 customers in every 100 leave each month.
• Direct costs take 20% of MRR. That covers hosting, payment fees and support tools.
• Team, by gross annual salary. The two founders start in month 1 at €30,000 each. A developer joins in month 3 at €36,000, the salesperson in month 13 at €30,000, a second developer in month 19 at €38,000 and a customer success hire in month 25 at €28,000. The employer pays gross × 1.4.
• Everything else costs €2,500 a month: €900 for the accountant and payroll, €400 for software and €1,200 for coworking.

Month	Customers	MRR	Total costs	Net burn	Cash at month end	Runway
1	6	€414	€11,583	€11,169	€388,831	34.8 months
6	33	€2,277	€16,155	€13,878	€320,825	23.1 months
12	83	€5,727	€19,845	€14,118	€229,435	16.3 months
18	153	€10,557	€24,311	€13,754	€136,918	10.0 months
24	216	€14,904	€29,614	€14,710	€40,213	2.7 months

Pick any row and follow it left to right. The drivers make the customers, the customers make the MRR, and MRR minus costs gives the burn, which is what eats the cash. Revenue grows every single month, yet the money runs out in month 27.

Scenario	CAC	Monthly churn	Salesperson closes	Cash runs out
Downside	€1,000	5%	3 a month	Month 24
Base	€700	3%	6 a month	Month 27
Upside	€550	2%	8 a month	Month 32

Two lessons come out of this. The scenarios move the date by only a few months, because at this stage the hiring plan drives the burn far more than revenue does: the decision with the biggest effect on runway is when you hire. And this founder should start raising around month 18 in the base case, earlier if the downside starts to look real. Being able to say that sentence, with the sheet behind it, is the whole point of the exercise.

The mistakes investors spot first

• A hockey stick with no drivers. Somebody typed 20% growth a month into a cell, or revenue triples in year two because that’s what the cell says. Every customer should come from a line you can explain.
• Hiring at gross salary. An employee in Italy costs the company well above gross pay. Eurostat’s latest figures put non-wage costs, mostly employer social contributions, at 28.1% of total labour costs in 2025. The EU average is 24.8%. Turn that around and €100 of gross pay costs about €139 on average. The TFR, a severance fund that grows each year by annual pay divided by 13.5 (roughly 7.4%), is part of that gap. Your exact figure depends on contract, sector and incentives, so ask a payroll consultant (consulente del lavoro) and use their number, not ours.
• Forgetting the founders. If the founders work in the company, their own social contributions are a cost too, even without a salary.
• Revenue confused with cash. An invoice isn’t money in the bank. Model payment terms and annual prepayments separately.
• No churn, or churn bolted on at the end. A subscription business that never loses a customer doesn’t exist.
• Forty tabs and false precision. Revenue to the euro in month 36 tells an investor you haven’t separated what you know from what you’re guessing.
• Numbers that disagree with the deck. Say slide 12 shows €1.2M of ARR and the model says €900,000. Now the meeting is a debate about which one is right, and it isn’t about your business any more.

How investors read a financial model in ten minutes

In a first meeting, nobody goes through your formulas cell by cell. They skim. So make these six places easy to find.

1. The assumptions sheet: are the inputs separate, visible and plausible?
2. The revenue build: does growth come from budget, CAC, conversion and sales capacity, or from a percentage someone typed in?
3. The hiring plan: who joins when, at what cost, employer contributions included.
4. The cash low point: the month the money runs out, and whether the round you’re asking for covers about two years.
5. The downside: what happens if CAC is 40% higher and churn doubles?
6. The unit economics: do CAC, churn and margin add up to customers who are worth more than they cost? Our guide to unit economics has the formulas.

After that, a careful investor compares the model with the metrics investors want at seed and with your pre-revenue valuation logic, because the amount you raise, the price and the runway have to tell one story. When you’re ready to raise, the guide to raising capital in Italy is the next stop.

Your checklist before you send it

• Every input on one assumptions sheet, coloured differently from formulas.
• Customers built from drivers, with churn from month one.
• Each hire listed with a start month, a gross salary and the employer cost.
• Thirty-six monthly columns, plus a one-page summary by year.
• Three scenarios, and the month cash runs out in each.
• The same figures as the deck, checked line by line.
• A short notes tab saying where each assumption comes from.