Getting pre-seed funding when you only have a pitch: the financial model that proves you understand your own business
Pre-seed money funds companies before they have a finished product or revenue. Every revenue projection you type is fiction, but investors read the financial model to verify you understand the business you intend to build. Six weeks out, you have a pitch deck, a story, and a spreadsheet you have been avoiding. The story wins the initial meeting, while the model survives the follow-up email when a micro-VC asks for the numbers. It provides the only proof of your thinking a spreadsheet can carry.
Why bother building a financial model when my product doesn’t exist yet?
A pre-seed financial model proves your grasp of unit economics and available levers. The numbers display your assumptions and reveal how you process information. Sloppy logic indicates you do not understand the mechanics of your business. Before your product ships, an angel verifies whether your customer acquisition cost, gross margin, and burn rate align. If you claim a $50 CAC and a $9/month price, and your model shows payback inside two months, an investor will test that logic against market reality.
“According to Forecastr, a pre-seed financial model focuses heavily on short-term cash management and demonstrating an 18-to-24-month runway to reach key operational milestones.”
Building a model with fake revenue numbers forces you to construct a logic engine. You define the rules that turn an assumption into cash out the door. While the revenue figures are placeholders, the structure remains real. Short-term financials show you can hit milestones without running out of money before the next round. The model validates that you understand what drives your cash burn.
Don’t investors spend less than a minute on the finance slide anyway?
The slide gets a brief glance during the pitch, but the underlying spreadsheet undergoes inspection during diligence. Angels and micro-VCs test whether your CAC and burn rate assumptions hold together. The widely cited 40-second average per slide measures the first pass, where an investor decides in under three minutes whether to take a meeting. After the meeting, an interested firm asks for the model.
“Investors spend an average of just 40 seconds analyzing the finance slide during their initial review of a pitch deck, according to Built In.”
A micro-VC will click on your CAC cell and ask for its source. If the answer is a hardcoded guess, the conversation cools. If the answer is a linked assumption you can defend, you keep the momentum. The finance slide advertises the metrics, and the full model provides the mathematical proof.
Can’t I just download a free financial model template online?
Marketplace templates often fail for AI startups because they assume linear enterprise sales (a rep, a quota, a contract) instead of usage-based pricing, GPU infrastructure costs, and self-serve funnel drop-off. A downloaded template relies on standard SaaS assumptions. The revenue tab expects seat-based pricing while you charge per 1,000 tokens processed. The cost tab lists hosting as a flat line while your variable cost is GPU inference. The sales tab models a quota-carrying rep while your growth comes from a self-serve funnel where 60% of sign-ups never convert.
“According to Carta, standard pre-seed rounds are predominantly raised using simple SAFEs or convertible notes rather than complex priced equity, making 12-to-18-month cash burn projections far more critical than capitalization tables.”
The primary risk is a hidden formula two tabs deep that assumes your cost of goods sold is 20% of revenue. That assumption fails when every active user costs you real compute. Presenting a template disconnected from how AI startups spend money signals a lack of understanding regarding unit economics.
Moddelix builds the driver logic for usage-based and infrastructure-heavy models, scaling GPU cost directly off usage. You input the assumptions, and Moddelix maintains the underlying formulas, avoiding the hardcoded errors of generic templates.
I am a solo technical founder. How do I build this without a finance background?
Map the financial model to the technical metrics you already track as an engineer. Treat your headcount, API calls, and GPU costs as the primary drivers of cash burn. Pre-seed investors prioritize business acumen over exact financial forecasting, so the burden rests on your understanding of the inputs.
“Founders raising pre-seed capital must prove plausible unit economics where 100% of customer monetization ultimately exceeds acquisition and delivery costs, according to Feel the Boot.”
A pre-seed model for an AI startup requires exactly four tabs.
The 4-tab minimalist pre-seed model:
| Tab | What it holds | The one question it answers |
|---|---|---|
| 1. Assumptions | Every driver in one place: sign-ups/week, conversion %, price, CAC, GPU cost per active user, salaries | “Where does every number come from?” |
| 2. Revenue | Usage-based build-up: active users × usage × price | “How do you generate revenue?” |
| 3. Costs & Burn | Salaries + infrastructure (GPU/API scaling with usage) + fixed overhead | “What does it cost to operate each month?” |
| 4. Cash & Runway | Opening cash minus monthly burn, month by month, until zero | “When does the cash run out, and does the raise reach the next milestone?” |
The model omits a balance sheet, GAAP accrual schedules, and deferred revenue waterfalls. Pre-seed investors do not expect these schedules. Every number on tabs 2 through 4 must trace back to a single cell on Tab 1. Avoid burying hardcoded values in a formula. When an investor clicks a cell, they should land on an assumption you can defend out loud. Traceable logic ensures you can explain the entire model on a call.
The financial model provides the quantitative proof
Before the product exists, the financial model provides the primary quantitative proof that you understand your own unit economics. Build the model so every cell traces to a verifiable assumption. You can construct those four tabs manually, mapping every cell to an assumption you can defend. Alternatively, you can build the same model in Moddelix.
Moddelix includes the driver logic and traceability by default, allowing you to focus on the pitch deck. Most founders start with a basic structure, then move into Moddelix when the follow-up diligence email arrives. The requirement is a single model where every number has a source, and Moddelix accelerates that process.
FAQ
Do I need a financial model for pre-seed funding if my product doesn’t exist yet?
Yes. At pre-seed, the financial model isn’t a revenue forecast — it’s proof you understand your own unit economics. Investors know early numbers are fiction; they read the model to test whether your assumptions about CAC, burn, and margin hold together logically. A coherent model shows you understand the business you’re raising for.
How long do pre-seed investors actually spend on the finance slide?
About 40 seconds during the first pass, out of roughly 4 minutes 10 seconds total on the deck, per DocSend data cited by Built In. But the slide only wins the meeting. The full spreadsheet gets inspected later during diligence, where investors click into your CAC and burn assumptions to see if they hold up.
Can I use a free financial model template for an AI startup?
Only carefully. Most marketplace templates assume linear enterprise sales — seat licenses, quota-carrying reps, flat hosting costs. AI startups run on usage-based pricing, GPU and API infrastructure that scales with usage, and self-serve funnels with heavy drop-off. A generic template breaks at exactly the assumption an investor scrutinizes most.
How many tabs should a pre-seed financial model have?
Four: Assumptions (every driver in one place), Revenue (usage-based build-up), Costs & Burn (salaries plus infrastructure), and Cash & Runway (month-by-month until zero). Skip the balance sheet and GAAP accrual schedules — at pre-seed, those answer questions nobody is asking yet. Every number on the later tabs should trace back to one cell on the Assumptions tab.
Can a solo technical founder build a financial model without a finance background?
Yes. Map the model to metrics you already track: headcount, API calls, and server or GPU costs as the drivers of cash burn. A pre-seed model is a system where the output is cash instead of latency. Pre-seed investors prioritize business acumen over accounting precision, so the burden is understanding, not certification.
