Assessing risk in early-stage companies: an investor's checklist

Institutional diligence works by reduction: you examine what exists — accounts, contracts, cohorts, churn — until the residual uncertainty is small enough to price. Early-stage diligence cannot work that way, because most of what you would examine has not happened yet. Applying a late-stage checklist to a young company produces a thick folder and no better decision.

What works instead is triage. You are not trying to eliminate uncertainty; you are trying to identify which uncertainties would be fatal, and then find the cheapest available evidence on those. This is the same discipline a venture builder applies during evaluation, turned around to face the investor's side of the table.

Six lenses

Ask for evidence, not plans

A business plan is a hypothesis with formatting. The useful question in every conversation is some version of: what evidence do you have, and how did you get it? Three specific customer conversations with named organisations tell you more than a market-size slide sourced from an analyst report. A paying pilot tells you more than a hundred waitlist signups.

• Who has paid, or committed to pay, and what did they buy?
• What did you believe six months ago that you no longer believe, and what changed your mind?
• Which assumption, if wrong, kills this business — and how are you testing it?
• What did the last three months cost, and what did they produce?
• Who has said no, and why?

That last question is the most informative and the least often asked. Founders who can describe their losses accurately have been in the market. Founders who cannot recall any have not.

The best question in the room

Rather than debating whether a forecast is achievable, invert it. Take the outcome the company is aiming at and work backwards: what would have to be true for this to work? Usually it resolves to three or four conditions — a certain conversion rate, a certain contract size, a channel that repeats, a hire the company has not yet made.

Now assess those conditions individually. Some will be plainly reasonable. One or two will be doing all the work, and that is where your attention and your questions belong. This technique does more to expose a weak case in twenty minutes than a week of document review.

Structural risk gets overlooked

Commercial risk is interesting, so people spend their time there. Structural risk is boring, so it gets a glance — and it is the category most likely to produce an avoidable problem. Check the basics properly:

• Cap table: who owns what, including anyone no longer involved. Dormant founders holding large stakes are a real obstacle to future rounds.
• Share class and rights: what you are buying and where it ranks. Note that SEIS and EIS require full-risk ordinary shares with no preferential rights on a winding up — see SEIS and EIS explained for investors.
• IP: is it owned by the company, including work done by contractors and by founders before incorporation?
• Key dependencies: a single customer, a single supplier, a single engineer who understands the system.
• Governance: who makes decisions, who is on the board, what reporting will you actually receive?

Valuation, honestly

Early-stage valuation is not a calculation; it is a negotiated position influenced by comparable rounds, competition for the deal and how much the founders need. Discounted cash flow models on a pre-revenue company are theatre. What is worth doing is a simple sanity check: at this entry price, what would the company need to be worth for this to be a materially good outcome, and is that plausible in this market?

Also model dilution honestly. If the company will need two or three further rounds, your position at exit is a fraction of your position today. A modest entry valuation is worth less than it appears if the company is likely to raise heavily.

Applying this to a venture-built company

When the company was created by a venture builder, several lenses answer differently — usually better, and occasionally in ways that need a second question.

The point of the model, from an investor's perspective, is set out in why invest in venture-built companies. The point of this article is that the model does not replace your own judgement — it just means your questions can be sharper and more specific.

Once you are comfortable assessing a single company, the harder problem is deciding how many to hold and at what size, which we cover in building an early-stage investment portfolio.