How to Build Your First Financial Model Without a Finance Background

Most founders avoid financial modeling because they think it requires an MBA. It doesn't. Here's how to build a model that's useful, defensible, and yours.

Published 2026-05-18 · 10 min read

Ask a non-finance founder why they don't have a financial model and you'll usually get one of three answers: "I'm not really a numbers person," "it's too early to project anything," or "my accountant said we don't need one yet." All three are wrong, but the third one is the most expensive - because it suggests the founder thinks the model is a deliverable for someone else.

A financial model is not a deliverable. It's a tool for the founder to think with, written in numbers so the founder can be honest with themselves. The MBA version of the model that shows up in fundraising decks is a downstream artifact of the simple, founder-built version that actually runs the company.

Here's how to build the founder version - no finance background required.

What the Model Is Actually For

A startup financial model has three real jobs at pre-seed and seed stage:

Tell you how long you survive. At current burn and current revenue, what is your runway? When does the company go to zero? This is the only number that matters in the short run.
Force you to write your assumptions down. A model converts vague optimism ("we'll grow fast") into specific numbers ("we'll add 12 customers a month at $200 ARPU") that you can defend or revise.
Let thoughtful readers disagree with you specifically. Investors, advisors, and co-founders can poke at a number. They can't poke at a hunch. The value of the model isn't in being right - it's in being wrong in a way that's productive.

What the model is not for, at this stage:

Predicting the future. No early-stage model has ever been right. That's not the goal.
Producing big numbers for an investor pitch. Investors discount your projections by 50–80% reflexively. Optimistic models lose credibility, not raise it.
Replacing your accountant. The model is forward-looking; the accountant looks at what already happened.

The Three-Layer Structure

A founder-built model has three layers, each built on the one below it. Build them in order.

Layer 1: Revenue

Revenue is what you charge, multiplied by who pays you, summed monthly. Most founders overcomplicate this. At pre-seed and seed, your revenue model has at most four moving parts:

New customers added per month (driven by your acquisition channels)
Average revenue per customer per month (ARPU)
Churn rate (the percentage of customers who leave each month)
Existing customer count (the running total)

The math is:

Customers this month = Customers last month × (1 − churn rate) + new customers added

Revenue this month = Customers this month × ARPU

That's it. Two formulas, repeated across 24 columns (one per month). Don't add expansion revenue, multi-tier pricing, contract-length adjustments, or annual prepay logic yet. You will get all of those wrong because you don't have the data to populate them. The point of the model at this stage is to be defensible, not detailed.

A worked example: assume you start January with 5 paying customers at $200/month each. You believe you can add 4 new customers per month (defensible: this is roughly one a week, which matches your discovery interview close rate). Churn is 5% monthly (defensible: B2B SaaS pre-PMF typically runs 5–8%).

Month	Starting	Lost (5%)	Added	Ending	Revenue
Jan	5	0	4	9	$1,800
Feb	9	0	4	13	$2,600
Mar	13	1	4	16	$3,200
Apr	16	1	4	19	$3,800

The numbers look small. That's correct. Pre-PMF revenue is small. A model that shows you at $50K MRR by month 6 with no specific reason for those numbers is fiction, and any investor you show it to will know.

Layer 2: Costs

Costs come in two groups, in order of importance: people, then everything else.

People costs are usually 70–85% of a pre-seed startup's burn. The line items are:

Each founder's salary (yes, including yours - set it at a realistic minimum, even if you're not paying it yet, so the model reflects the true cost of the team)
Each employee or full-time contractor
Payroll tax, benefits, and overhead - budget roughly 20–25% on top of base salary for U.S. employees

That's the people layer. Sum them per month. Now you have your payroll line.

Everything else is a short list at pre-seed:

Software / SaaS spend (your tooling stack)
Hosting and infrastructure (cloud, APIs, third-party services)
Marketing spend (paid ads, content, tools)
Legal and accounting (one-time and recurring)
Office, equipment, travel
Buffer (10–15% on top, because you will forget something)

Sum everything. That's your monthly burn. For most pre-seed startups, monthly burn lands between $15K and $80K, depending on team size and salary levels. If yours is meaningfully outside that range, double-check your assumptions before going further.

Layer 3: Cash

Cash is the only line that actually decides whether you exist next quarter. Build it last, but watch it most.

The formula is simple:

Cash at end of month = Cash at start of month + Revenue − Costs

That's it. Walk that forward, month by month, and the month where the cash line crosses zero is your default-dead date.

A model with $300K in starting cash, $4K monthly revenue growing slowly, and $35K monthly burn will show cash dropping steadily until somewhere around month 10 or 11, the line crosses zero, and the company is dead. That date is the most important number in your entire spreadsheet - more important than any growth rate, any LTV:CAC ratio, any ARR projection.

You should know this date the same way you know your home address. Re-check it every Monday.

The Three Scenarios Every Model Should Have

A single-column projection (the "plan") is brittle. Every founder model should have three scenarios on the same sheet:

Base case: what you actually believe, based on your best read of evidence today. This is not your aspiration - it's your honest estimate.
Downside case: what happens if your two or three most important assumptions land worse than you think. Half the new-customer rate. Twice the churn. A hiring delay.
Upside case: what happens if everything goes right. You hit your acquisition target, churn improves, you raise the price.

The point of the three scenarios is not to show all three to an investor - base case is usually the only one that leaves the spreadsheet. The point is to see your own range. If your downside case still gives you 12 months of runway, you can take more risk. If your base case is six months and your downside is three, you're already on the back foot and you need to either cut burn or raise sooner than you planned.

What to Defend, What to Just Label as a Guess

Every input in your model is either defensible (you have evidence) or assumed (you're guessing). The model should distinguish these visibly - highlight, comment, color - so a reader (and you, six weeks from now) can tell them apart.

Defensible	Assumed
Current MRR (the bank statement says so)	Year-2 expansion revenue (you have no expansion data yet)
Salary commitments (signed offer letters)	Marketing CAC at scale (no scaled marketing yet)
Current burn (last 3 months of actuals)	Series A timing and size
Conversion rate from your last 50 demos	Conversion rate at 10× current traffic

Defensible inputs anchor the near-term runway numbers. Assumed inputs drive the longer-term lines. The closer to month 1 of your projection, the more defensible the inputs should be. The further out, the more they're guesses - which is fine, as long as you label them. The danger isn't having guesses in the model. The danger is forgetting which lines are guesses.

The Two Numbers a Founder Should Know From Memory

A founder running a model well can answer two questions instantly, without opening the spreadsheet:

What is your current runway in months? (At today's burn, today's cash, today's revenue.)
What is your default-dead date? (The specific calendar month when cash hits zero if nothing changes.)

If you can't answer those two questions right now, your model is not running your company. It's an artifact you built once and stopped looking at - which is the failure mode for 80% of founder-built models.

Update It Monthly. Don't Build a New One.

The most common founder mistake with financial models is building a new one every six months when the old one is so far from reality that it's embarrassing. The cost of this is high: every rebuild loses the audit trail of your assumptions, hides the gap between what you predicted and what happened, and wastes a week of founder time.

The discipline that pays: update the model on the first business day of every month. Drop the actuals from last month into the model. Compare actuals vs. base-case forecast. Where they diverge significantly, ask whether you need to update the assumption, change the strategy, or both.

This monthly ritual is the most valuable hour you'll spend in any month as a founder. It's the moment where optimism collides with bank balance, and the company adjusts. Founders who do this routinely don't run out of money by surprise. Founders who don't, do.

What This Looks Like in 1tab.ai

1tab.ai includes a Financials module designed for founders without a finance background - pre-built revenue, cost, and cash layers, three-scenario projections, automatic actuals-vs-forecast tracking from your CRM and bank syncs, and a runway widget that lives next to your tasks and OKRs so the default-dead date is never out of sight.

Build a model you actually trust →

← Back to Blog