How should a founder prepare for an expert interview?

An expert is anyone who has spent years inside the problem: a former operator, a domain specialist, someone who ran the category at a competitor. Talk to them before customers, in order of how close they are to the problem. Prepare so you never ask what you could have read. Ask for help, never for validation, and use the end of the conversation to get the next one.

Who counts as an expert for startup research and in what order should you talk to them?

An expert is anyone who has spent years inside the problem: a former operator, a domain specialist, someone who ran the category at a competitor. Talk to them before customers, in order of how close they are to the problem. Prepare so you never ask what you could have read. Ask for help, never for validation, and use the end of the conversation to get the next one.

How do you get introductions to more experts and to customers?

An expert is anyone who has spent years inside the problem: a former operator, a domain specialist, someone who ran the category at a competitor. Talk to them before customers, in order of how close they are to the problem. Prepare so you never ask what you could have read. Ask for help, never for validation, and use the end of the conversation to get the next one.

Expert conversations · Product Sprint

§The question this chapter answers: who can compress years of the problem into an hour, and how do you earn that hour?

§Experts are the source founders under-use, for two reasons. The first is that they seem hard to reach. The second is that founders arrive wanting to be told the idea is good, and experts can smell that from the first sentence.

§ 3.1Who counts#

§Anyone whose years are in the problem rather than in your solution. In rough order of value:

§Former operators. People who did the job your customer does, or ran the business your customer runs, and have left. They know the problem intimately and no longer have a reason to be polite about it.

§Domain specialists. Analysts, researchers, consultants, regulators, the person who wrote the report you read yesterday. They know the shape of the market and where the bodies are buried.

§Category veterans. People who built or sold a in the space, including at competitors. They know which of your assumptions their company already tested and what it found.

§Current operators. The person doing the job today. They are also your future customer, and the line between expert conversation and blurs here. Treat them as an expert about the domain and as a customer about their own behavior.

§The order is by distance from the problem and freedom to speak. A former operator can tell you the thing a current one cannot.

§ 3.2Preparation, or why most expert calls are wasted#

§You get one hour, at most, and possibly one. Anything you ask that you could have read is a minute of that hour thrown away, and the expert knows it. is the preparation. Arrive knowing the market, the competitors, the vocabulary and the failures, and spend the hour on the questions that remained silent.

§Write five questions. Rank them. You will get through three. The fourth and fifth are there so you know what to drop.

§ 3.3Ask for help, not validation#

§Founders open expert conversations with a pitch. It is the wrong opening for two reasons. It spends the hour on your talking rather than theirs. And it puts the expert in the position of judge, which produces a verdict, usually kind, usually useless.

§Open instead with the problem and with your ignorance. "I am trying to understand why independent restaurants still count inventory by hand. You ran forty of them. What am I missing?" That question gets a story. Stories contain the facts you came for and the ones you did not know to ask about.

§Then listen. Steve Portigal's rule for user interviews applies doubly here: the interview is about them, and every sentence you speak is one they do not. When the expert says something surprising, follow it. When they say something you expected, move on. The surprises are the whole yield.

§ 3.4The last fifteen minutes#

§The end of an expert conversation is where its compounding value lives. Three asks, in order.

§"Who else should I talk to?" Every expert knows two more. Ask for one introduction they would actually make, not a list of names. Rob Fitzpatrick's test for a real holds: an introduction offered is worth ten names given.

§"Who has this problem worst?" Experts know where the pain concentrates. This is how you find the customers for the conversations in chapter 5, and it is usually a narrower group than the one in your research plan.

§"May I come back?" Ask whether they would look at what you build. Most say yes, and some mean it. Those become the informal advisors every early company runs on.

§ 3.5What you leave with#

§One page per expert, written the same day: three things you did not know, one assumption they contradicted, the names they gave you, and what they agreed to. Then an updated research plan with the silent assumptions now marked, and a shortlist of who has the problem worst. That shortlist is your recruiting list for chapter 5.

3. Expert conversations