What Investors Look for in a Financial Model
By Tom McCollum, ACA · Former KPMG Corporate Finance · Compass & Ledger
The short answer
Investors don't expect your forecast to be right. They expect it to be believable and testable. They check that cash ties out, revenue is built from real drivers, gross margin includes every cost, the hiring plan matches the cost base, and that you have a downside case. Most then rebuild your key numbers in their own format, so anything they can't trace costs you credibility.
We build financial models for founders going out to raise, and we've sat on the investor side reviewing them. This is what an investor or their analyst does when they open your model, and where models most often fall down.
The first 10 minutes: what gets checked first
Most investors follow a similar sequence:
- Does the balance sheet balance, and does cash tie to the cash flow statement? A model that doesn't balance loses credibility immediately, however good the business.
- Where do the inputs live? They look for an assumptions tab. If key numbers are typed directly into formulas, they can't test them quickly.
- What drives revenue? They trace the top line back to its drivers.
- When does the cash run out? Runway under the base case, and under a downside.
Get these four right and the rest of the conversation is about your business, not your spreadsheet.
Revenue: drivers, not growth rates
"Revenue grows 15% a month" isn't a forecast. Investors want to see revenue built from what actually produces it:
- SaaS: new customers × price, plus expansion, minus churn, often by cohort.
- Marketplace or e-commerce: traffic × conversion × order value × repeat rate.
- Services or B2B: pipeline × win rate × deal size, with sales-cycle timing.
The test they apply is simple: if marketing spend doubles or a salesperson starts late, does revenue move sensibly? If revenue doesn't respond to the things that drive it, the model isn't doing its job.
Gross margin: the most restated number
Gross margin is where we most often see numbers restated during due diligence. The usual cause is costs of delivering the product left out: hosting and infrastructure, fulfilment and shipping, payment processing fees, customer support, implementation. If a cost goes up when you sell more, it probably belongs in cost of sales.
Investors also look for margin to improve for a stated reason, such as volume pricing from suppliers or lower hosting costs per customer, not simply because the model says so.
Unit economics
Expect to be asked for, and to defend:
- Customer acquisition cost, by channel where you can.
- Payback period: months of gross profit needed to recover acquisition cost.
- Retention: logo and revenue retention, ideally by cohort.
- Lifetime value, with the assumptions behind it.
- For SaaS, the burn multiple: net burn divided by net new ARR. It shows how efficiently you turn cash into growth.
The numbers matter less than whether they're consistent with the rest of the model. A payback of six months alongside marketing costs that imply eighteen is a red flag.
The hiring plan and cost base
People are usually the biggest cost, and investors test the hiring plan hard:
- Is every hire linked to a reason, such as revenue capacity, product milestones or support load?
- Are start dates realistic, with recruiting time built in?
- Do new salespeople ramp up gradually, or produce full revenue from day one? Day-one productivity is one of the most common modelling errors.
- Are salaries fully loaded, including payroll taxes, benefits and equipment?
Cash: working capital, tax and timing
Profit isn't cash, and investors will check that your model knows the difference:
- When customers pay versus when revenue is recognised. Annual upfront billing helps cash; 60-day payment terms hurt it.
- Inventory and supplier payment terms, for anything physical.
- Taxes, including payroll and sales tax.
- The timing of the raise itself: cash from the round shouldn't land the day the model starts.
Scenarios
Show at least a base case and a downside. The downside isn't there to show you're pessimistic; it shows you know which assumptions matter most and what you'd do if they slip. The question investors most want answered is: how long does the cash last if growth is slower, and what would you cut?
Common mistakes that cost credibility
- Hardcoded numbers inside formulas.
- A balance sheet that doesn't balance, or a check line that's been hidden.
- Salespeople productive from day one.
- Gross margin that improves with no explanation.
- Deck figures that don't match the model.
- Five-year hockey sticks with nothing in years one and two to support them.
- A model so complex only its author can use it.
How detailed should it be?
Seed: monthly for 24 months, a clear hiring plan, burn and runway. Pre-revenue companies still need one; it shows you understand your cost base. Series A: add cohorts, unit economics by channel and a fuller three-statement build. The right model is the simplest one that answers every question an investor will ask.
Frequently asked questions
Do investors expect a three-statement model at seed?
Many seed investors focus on the profit and loss, burn and runway, but a three-statement model that ties shows rigour and is expected by Series A.
How many years should the forecast cover?
Typically three to five years, monthly for at least the first 24 months.
Should the model be in Excel or Google Sheets?
Either. What matters is that an analyst can open it, find the assumptions and change them.
Need an investor-grade model?
We build three-statement, driver-based models as part of Investor Ready in 30 Days, or on their own from $4,000.
Thirty minutes. No pitch. Your situation, and what we would do about it.
If we are a fit, we will say so. If we are not, you will hear that too, and where to look instead. Either way you leave with a clearer picture of what good finance looks like for your business.