Your engineers report that the personalised feed would need a response time the current data pipeline cannot reach, and that rebuilding the pipeline is two quarters of work. Which risk has that answered, and what does it say about the rest?

The principle — the four risks are four questions, and an answer to one is an answer to one.

Why the key is correct — the engineers' report is exactly the evidence feasibility asks for, since that risk is whether our engineers can build what we need with the time, skills and technology we have A pipeline that cannot reach the required response time inside the available quarters is a direct answer.

Why the others are wrong — treating the risks as a sequence adds an order Cagan never gives them, and a team routinely learns about the build before it learns anything about demand. Letting a hard build lower confidence in the idea imports one answer into three questions that were never asked. And reading it as value risk describes something that has not happened, because nothing has reached a customer.

Remember this — one risk answered is one risk answered; the other three are exactly where you left them.

Sources — Marty Cagan on the four big risks, Silicon Valley Product Group.