The seed financial model 2026: a drivers-based template

The seed financial model 2026 VCs accept is small. Five to seven sheets, three to five defendable assumptions, a 36-month horizon, base/bear/bull toggles. The job is not precision, it is showing you understand the levers that drive the next round. This guide is the exact sheet list, the drivers to defend, and the seven things to cut.

Contents

• What a VC-ready financial model actually looks like at seed
• The seven sheets every founder financial model template needs
• The five drivers you must defend in a SaaS seed model
• Inside the seed-stage three-statement: what stays, what goes
• The 36-month horizon and base/bear/bull scenarios
• What to leave out of your drivers-based financial model
• How VCs read your seed financial model in diligence
• FAQ

Most seed founders ship one of two unfundable models. The over-engineered one has fifty tabs, cohort retention curves at zero customers, and revenue projections to four decimal places. The under-engineered one is a single sheet where Revenue × 0.1 = Profit. Neither gets a check.

What VCs actually read at the seed financial model 2026 stage sits between those failure modes. Five to seven sheets, three to five drivers you defend out loud, thirty-six months, base/bear/bull. The point is not to predict the future; it is to prove you understand which numbers, if wrong, kill the company.

This guide is the exact sheet architecture, the drivers worth defending in a SaaS seed model, the cuts that signal seniority, and what partners actually look at when your model lands in their inbox.

What a VC-ready financial model actually looks like at seed

A VC-ready financial model at seed is a tool for making three decisions: who you hire and when, how long the round lasts, and what the next round looks like.

Everything else is decoration. A partner has thirty decks open this week and will spend twelve to twenty minutes inside your model. They check three things: whether the hiring plan is real, whether the unit economics are coherent, and whether the cash runway clears the Series A milestone.

The 2026 environment rewards capital efficiency. The share of sub-$5M VC deals fell to 50.3% in 2025, a decade low (PitchBook-NVCA Venture Monitor Q3 2025). Smaller rounds are scarcer, and the founders winning them ship leaner models with sharper logic.

The seven sheets every founder financial model template needs

The founder financial model template that survives diligence has exactly seven sheets. Not five, not twelve. Each does one job. If a sheet's purpose cannot be stated in one sentence, delete it.

1. Assumptions. All driver inputs live here and only here: conversion rate, ACV, sales-cycle length, gross margin, churn, salary bands, hiring-plan growth rate. Every other sheet pulls from this one. A partner who wants to flex conversion changes one cell and watches the model react.
2. Hiring Plan. Monthly headcount by role with a hire date, a fully-loaded cost (base × ~1.33 for taxes, benefits, equipment, per Kruze Consulting's 2025 model guide), and a function tag. Headcount is the single largest cost driver in a seed startup.
3. Revenue. A monthly build from drivers to bookings, ARR, and recognized revenue. For SaaS: new logos × ACV minus churned ARR. For usage-based: paying customers × average revenue per account. The formula must be visible; partners back-solve from drivers.
4. Expenses. Monthly opex pulled from the hiring plan plus four fixed categories: tools, rent, marketing, contractors. That is it. No granular per-vendor breakdown.
5. P&L. Revenue minus COGS gives gross profit; gross profit minus opex gives operating income. Three rows per month. No deferred-revenue gymnastics, no D&A, no taxes.
6. Cash. Starting cash plus net income plus financing inflows minus capex. The runway line is read here, month by month.
7. Scenarios. Base, bear, and bull toggles. A single dropdown in Assumptions flexes conversion, hiring pace, and churn so all three cases populate from one switch.

Kruze's 2025 guide explicitly recommends skipping the balance sheet and the full cash flow statement at seed. Both belong to the Series A model and add complexity no partner reads.

The five drivers you must defend in a SaaS seed model

A SaaS seed model stands or falls on five numbers. Get these right and the rest of the model writes itself. Get them wrong and no amount of decoration saves it.

For context, public SaaS sits at 110% median net dollar retention, and sub-$1M ARR startups rebounded to a 100% median year-over-year growth rate in 2024 (High Alpha / OpenView 2024 SaaS Benchmarks). You do not need to hit those exactly; you need to know where you sit against them and why.

Defend each driver in one sentence. "Our ACV is $24k because that is what our last five paid pilots converted at, with two outliers excluded." That is a defense. "Our ACV is $50k because it grows 10% a year" is not. The first sentence shows you have data; the second shows you have a spreadsheet.

The trap most seed founders fall into is driver inflation. Modeling fifteen inputs you cannot defend signals that you do not know which three matter. Pick the five that determine whether the business works, prove you can pressure-test each, and leave the rest as constants.

Inside the seed-stage three-statement: what stays, what goes

The seed-stage three-statement at seed is really a one-and-a-half-statement: a full P&L and a cash position, with the balance sheet stubbed out until Series A.

Kruze's guidance is explicit: at seed, build the Income Statement and Cash Position; skip the Balance Sheet and the full Cash Flow Statement until Series A diligence requires them. The reason is signal-to-noise. A balance sheet at zero revenue is mostly cash and noise. A full cash flow statement is operating, investing, and financing flows where investing is zero, financing is one row (your SAFE), and operating is just net income.

What you keep:

• P&L by month: Revenue, COGS, Gross Profit, Opex (split into Personnel, Tools, Marketing, Contractors, Rent), Operating Income. Six rows, twelve to thirty-six columns.
• Cash position by month: Starting Cash, Net Income, Financing Inflows, Capex, Ending Cash. Four rows, twelve to thirty-six columns. Runway is read off this sheet.

What you cut:

• Balance Sheet: Stub it. A single cell where cash equals ending cash on the Cash sheet is enough.
• Cash Flow Statement (full GAAP): Skip. The cash-position sheet does the job.
• Working capital schedule: Skip. Receivables and payables at seed are negligible.

This is the difference between a model that signals founder-built and pragmatic versus one that signals CFO-template-padded. Partners can tell which is which in thirty seconds.

The 36-month horizon and base/bear/bull scenarios

A 36-month seed-stage projection is the standard. Not twelve, not sixty. The reason is mechanical: you raise 18-24 months of cash at seed and need to show what the company looks like at the next round, plus a buffer (Kruze).

Twelve months is too short to show the Series A story. Sixty months is fantasy at seed; nobody believes month 47. Thirty-six months gives you the seed-funded period (months 0-18), the next-round timing (months 18-24), and the year after (months 24-36) where partners check whether the business compounds or stalls.

Three scenarios, one toggle. The scenario dropdown sits in Assumptions and flexes the same three drivers across all three cases.

Bear case is the one partners actually study. The question they answer in their head: how long do you survive if the world does not cooperate, and what would you cut to extend that? A bear case that runs out of cash four months before the next round is a signal to trim the hiring plan, not a reason to walk. A model with no bear case at all is a reason to walk.

Kruze recommends stress-testing the model by varying conversion and growth rates explicitly: scenarios are not decoration, they are the test of whether your drivers are connected to outputs at all.

A model with no bear case is not a model. It is a forecast you do not believe in.

What to leave out of your drivers-based financial model

The fastest way to look unserious is to over-model. Every line below shows up in rejected seed models. Cut all of them.

• Cohort retention curves before you have cohorts: A curve drawn from three customers is noise. Under twenty paying customers, retention is a single number (gross logo retention this month) and a one-line note, not a curve.
• Multi-currency translation: If you have one European customer, charge them in USD or hold FX as a constant. A multi-currency hedging schedule at seed is a tell.
• Complex equity waterfalls: The cap table belongs in Carta, not in the model. A simple post-money table on Assumptions is enough: founder shares, ESOP, SAFE conversion at cap.
• Granular per-SKU COGS: A blended gross margin with a footnote on why it lands where it does beats a per-component breakdown of compute, storage, payment processing, and human review. Save it for Series A diligence.
• Full balance-sheet forecasting: Already covered. A line for ending cash is the balance sheet at seed.
• Marketing ROI / CAC payback models: At seed you do not have statistically meaningful CAC payback data. Model marketing as a fixed monthly category with a one-line assumption on channel mix.
• Tax provisioning: You are not profitable. The tax line is zero. Stop modeling state nexus.

This is the what NOT to include test almost no other guide on seed models writes down. Every item above is a Series A artifact dropped into a seed model by a founder who copied a template they did not understand. Partners notice immediately; it is one of the cheapest disqualifiers in diligence.

How VCs read your seed financial model in diligence

Partners read the seed financial model 2026 in three passes, in this order.

Pass one (two minutes): the hiring plan. This is the first sheet they open. Over 80% of seed-startup spending goes to payroll, rent, and contractors, with payroll dominant (Kruze). The hiring plan IS the model. Partners read the next twelve hires, the function mix, the timing, and the loaded cost (base × ~1.33), and form an opinion of whether you understand the team you need.

Pass two (five minutes): the drivers. They jump to Assumptions and read conversion rate, ACV, gross margin, churn, and growth rate. They are checking whether the numbers are defendable and whether changing one cell flexes the whole model. A hard-coded revenue row that does not move when you change conversion is a kill signal.

Pass three (ten minutes): the runway and the bear case. They open Cash, find the month you run out, and check whether it matches the round size you are asking for. Then they flip the scenario dropdown to bear and watch what happens. A bear case identical to the base case means the scenarios are decoration.

One model, fifteen minutes, three decisions. Build the model to survive that read. Not exhaustive, not comprehensive, not impressive. Build it so the partner ends the read saying: "I see what they are betting on, I see what they would cut if forced to, and I can defend this internally."

The financial model is what lets your seed valuation 2026 thinking and your seed-stage traction metrics translate into a credible round size and runway story, and it is the spine of your VC due diligence checklist prep. Tools like Causo can pre-fill the drivers sheet from pipeline data, but for most founders the value is in writing the formulas yourself; the act of building it is the diligence prep. For a wider view of round structure and timing, see the guide on raising a seed round in 2026.

FAQ

What is a drivers-based financial model? A drivers-based financial model derives every output from a small set of input assumptions you can defend out loud. Instead of hard-coding revenue or burn, you build the model so that changing three to five drivers (conversion, ACV, hiring pace, gross margin, churn) flows through every line. The point is to make trade-offs visible: drop conversion 20% and you can see which month the runway breaks.

How many tabs should a startup financial model have? A seed-stage model needs five to seven sheets: Assumptions, Hiring Plan, Revenue, Expenses, P&L, Cash, and Scenarios. Kruze Consulting recommends explicitly omitting the balance sheet and full cash flow statement until Series A. More than seven sheets at seed usually signals padding, not rigor.

What should you NOT include in a seed financial model? Skip cohort retention curves before you have cohorts, multi-currency translation, complex equity waterfalls, granular per-SKU COGS, and full balance-sheet forecasting. These are Series A artifacts. At seed they signal a founder who copied a template they did not fully understand.

What goes into a 36-month seed stage projection? Three years of monthly hiring plan, monthly revenue built from a small set of drivers, monthly opex pulled from headcount plus four or five fixed categories, and a running cash balance. The 36-month horizon lets you show the Series A milestone, the gap to that milestone, and whether the round you are raising clears it with six months of buffer.

Do VCs care about financial models at seed stage? Yes, but not for the numbers themselves. Partners read the seed model to test whether you understand the unit economics of your business and the cost of the team you plan to hire. A model with thoughtful drivers and a credible hiring plan earns trust; a hockey-stick with round-number expenses costs you the meeting.

Related on the hub

• How to raise a seed round 2026: the end-to-end playbook — Related fundraising basics guide.
• Biotech seed fundraise 2026: platform vs asset, tranches, milestones — Related fundraising basics guide.
• AI founder seed 2026: what changed and the playbook that works — Related fundraising basics guide.