What Creandum, Profounders, and B2Ventures Actually Look For in Your Model
Tier-one VCs are not impressed by formatting or complexity. They are looking for internal consistency, sourced assumptions, honest scenarios, and a founder who can defend every number without looking at the screen. The model signals whether you understand your own business.
Author: Yanni Papoutsi · Fractional VP of Finance and Strategy for early-stage startups · Author, *Raise Ready*
Published: 2025-03-08 · Last updated: 2025-03-08
Reading time: \~8 min
Why the Model Is a Credibility Test
Before a VC writes a check, they are evaluating three things about the model: whether the numbers are internally consistent, whether the assumptions are defensible, and whether the founder presenting the model actually understands what drives it.
A beautiful model with circular logic fails the first test. A model with "industry average" as the only source for every assumption fails the second. A founder who cannot explain why the CAC assumption is $320 and not $500 fails the third.
The model is not just a financial document. It is a proxy for how the founder thinks about their business.
What experienced investors check first:
Revenue drivers Is revenue an output of real mechanics or an assumed growth rate?
Gross margin | Is it benchmarked and are COGS correctly separated from OpEx?
Headcount plan | Is it role-by-role with loaded costs or a lump-sum? Scenario | Are there at least three scenarios and does structure | conservative show survival?
Assumption | Does every key input have a cited benchmark or logic? sourcing
Cash flow timing Does the model show monthly cash, not just annual P&L?
The First Five Minutes of Model Diligence
Experienced investors develop a pattern for reviewing models quickly. Having been through this process with investors including Creandum and Profounders, the first five minutes typically cover:
Revenue tab: Can I trace where this number comes from? Is it customers times conversion times ACV? Or did someone type a growth rate? Gross margin: What is it, how does it compare to sector benchmarks, and is cost of goods sold separated from operating expenses? Headcount: Is there a named hire plan or a salary line? Named hires signal operational thinking. A salary line signals the founder has not thought this far.
Scenarios: Is there only one? Single-scenario models communicate either inexperience or that the founder has not stress-tested their own business.
The ask: Is the raise size connected to specific milestones? Or is it a number that feels right?
If any of these five checks fail, the rest of the model review becomes much more sceptical.
Key insight: The founders who raise at good terms are almost always the ones who can sit in a room without their laptop and explain why the model makes the assumptions it does. The model is the preparation. The meeting is the test.
What \"Sourced Assumptions\" Actually Means
"Sourced assumptions" does not mean there is a footnote somewhere. It means the investor can trace every material input to either: 1. Your own data (actuals from the last 3-6 months that justify the
assumption)
2. A named external benchmark (OpenView SaaS benchmarks, DocSend
fundraising data, a named comparable company)
3. A logical construction you can explain out loud
"We used industry average" is not a source. "Our current pilot has a $280 CAC across three channels, and we expect blended CAC to decrease to $220 by Month 12 as the referral channel scales, based on the referral share trajectory we observed in the first 90 days" is a source.
The difference is whether the assumption is tethered to something real or floating.
The Conservative Scenario Test
The single most revealing scenario in any model is the conservative case. What happens if revenue is 30% below base case? When does cash run out? What decisions would the team make?
Investors use this scenario not to predict failure but to assess whether the founder has thought about the business from the perspective of what can go wrong, not just what will go right.
A conservative scenario that still shows the company growing comfortably is not a conservative scenario. A conservative scenario that shows the company running out of cash in Month 14 reveals the actual risk profile of the raise.
Show it anyway. Founders who hide the conservative case lose credibility faster than those who show it and explain how they would respond.
Common Model Failures in VC Diligence
Revenue as a growth rate, not a driver calculation
COGS and OpEx mixed, producing a misleading gross margin
No monthly cash flow (annual P&L hides cash gaps)
Single scenario with no downside analysis
Headcount modelled as a total salary figure with no hire-by-hire
plan
Circular references that cause the balance sheet not to balance Key metrics (LTV, CAC, payback) calculated incorrectly or not shown
Frequently Asked Questions
How detailed should a pre-seed model be?
Detailed enough to show you understand the unit economics and burn profile. A pre-seed model should have monthly cash flow for at least 18 months, a driver-based revenue forecast (even if the drivers are estimates), a headcount plan, and a clear connection between the raise amount and the milestones it funds.
Do VCs share model feedback before a pass?
Sometimes. The best investor relationships include honest feedback. But most passes come without detailed model critique. If you receive feedback, treat it as valuable data even if the relationship did not proceed.
Should the model shown to investors be the same as the internal model?
The investor model should be a cleaned and documented version of the internal model, with an assumptions tab and summary. It should not be a separate document built specifically to look good. Inconsistency between the two creates risk if both are ever reviewed simultaneously during diligence.
Summary
The model VC investors review is a test of how well you understand your own business. Revenue must come from real drivers. Assumptions must be sourced. Scenarios must be honest. Monthly cash flow must be visible. And the founder must be able to explain all of it without the spreadsheet. These are not complexity requirements. They are clarity requirements. A simple model that passes all of these tests is stronger than a complex one that fails them.
Get the complete guide with all 16 chapters, exercises, and model templates.
Get Raise Ready - $9.99