A few different looks at the dimensions of risk in early-stage startups. Not a sufficient framework for evaluating an idea (from the founder or venture capitalist perspective), but a good start in categorizing major risk vectors. Startups are high-risk and have a corresponding risk-return profile to show for it. But classifying, mitigating, and thinking through the risks might be helpful.
Market risk — The only thing that matters (according to Marc Andreessen). If the market doesn’t exist, even the best teams can’t sell into it.
Even if the market exists (or will exist) — you can divide market risk into two different categories:
- Timing risk — Is the startup either (1) too early or (2) too late to the market? Often the accompanying question is, why now? I think the most common failure mode is “too early” — lots of entrepreneurs have mistakenly believed that they were “too late” when they were simply much deeper into the problem space than everyone else. The first-mover advantage is overrated.
- Total addressable market (TAM) risk— Sometimes, the market might be valid and timely, but the size isn’t big enough to support venture-sized returns. For the most part, software markets tend to be long-tailed when they exist.
Andy Rachleff puts it like this:
When a great team meets a lousy market, market wins.
When a lousy team meets a great market, market wins.
When a great team meets a great market, something special happens.
Execution risk — Can the team build and sell the product? Do they have the right skills to scale the company? Do they have access to the talent they need? Can the company navigate capital management, people management, and more? Will the company be able to execute on its goals to get to the next milestone?
Technical risk — A corollary of execution risk is technical risk — can the product be built? Exogenous variables can prevent even great teams from reaching their ambitious goals. Most important for companies that are dealing with hardware or cutting-edge research.
Platform Risk — Relevant in the era of foundational models and tech giants. Past examples of this failure mode are Zynga/Facebook, and all of the companies that built on Twitter’s API (before it was shut down for the first time). Current examples might be companies building AI products on foundational model providers (e.g., OpenAI) or open-source companies selling managed services on cloud providers (e.g. Elastic/AWS).
Business model risk — Less of an issue when it comes to pure software models, but I believe that it’s probably more relevant now that there’s a cloud and infrastructure tax (are you just reselling storage or compute?). You could probably put GTM strategies into this as well, although those are much easier to iterate early on with the right team.
Some more minor ones:
- Capital structure risk — Does the company have a capitalization table that makes it hard for future investors to invest (e.g., stacked SAFEs, deep preference stacks, and unwieldy terms)?
- Legal/Regulatory risk — particularly for crypto startups, the lack of a legal framework or the emergence of a new policy can be an existential threat to the business.