What diligence looks like on an early-stage deal

Diligence on a mature company is largely verification: the numbers exist and the question is whether they hold. On an early-stage company there is little to verify, so the work moves to judgement under uncertainty.

What can still be checked

Even at seed stage, several things are factual rather than speculative. Who founded the company and what they did before. Which customers pay, how much and since when. What the product does today, not in the roadmap. How the previous rounds were structured and who owns what.

• Customer references, including churned ones

• Cohort behaviour rather than aggregate growth

• Gross margin after support and hosting costs

• Cap table, option pool and existing preferences

• Founder track record and reference calls

Where judgement takes over

Market size is the part most often misused. A number taken from a consultancy report says little about the segment a company can realistically reach in the next three years. It is more useful to build the market from the bottom: how many customers exist, what they currently spend and on what.

The terms are part of the analysis

A reasonable company at the wrong terms is a bad investment. Preference structure, pro rata rights and the option pool all change the outcome for an investor without changing anything about the business.

The output of diligence is not a score. It is a written case that states what has to be true, what has been checked and what remains a bet.