What research should a startup do before talking to customers?

Sources of insight are not equal and their order matters. Secondary research first, because it is cheap and tells you who tried before. Experts next, because they compress years into an hour. Communities next, because that is where the language lives. Customers last, because by then you know what to ask. Before any of it, write down what you assume, who you think has the problem, and what would change your mind.

In what order should a founder use secondary research, experts, communities and customer interviews?

Sources of insight are not equal and their order matters. Secondary research first, because it is cheap and tells you who tried before. Experts next, because they compress years into an hour. Communities next, because that is where the language lives. Customers last, because by then you know what to ask. Before any of it, write down what you assume, who you think has the problem, and what would change your mind.

What goes in a research plan for a startup?

Sources of insight are not equal and their order matters. Secondary research first, because it is cheap and tells you who tried before. Experts next, because they compress years into an hour. Communities next, because that is where the language lives. Customers last, because by then you know what to ask. Before any of it, write down what you assume, who you think has the problem, and what would change your mind.

A hierarchy of sources · Product Sprint

§The question this chapter answers: where should you look first, and how will you know when you have seen enough?

§Most founders start with customer interviews because the advice says talk to customers, and end up asking strangers to explain a market the founder could have read about in an afternoon. The interview is the most expensive source you have. Spend it last.

§ 1.1The order and why it holds#

§Secondary research first. Reports, competitor sites, app store reviews, pricing pages, funding announcements, the post-mortems of companies that tried this before. It costs nothing but attention, and it answers the questions you should not waste a human on: how big, who else, what happened last time.

§Experts second. People who have spent years inside the problem: former operators, analysts, the person who ran the category at a competitor. One hour with one of them saves a month of interviews, because they tell you which of your questions is naive before you ask it of a customer.

§Communities third. Forums, subreddits, Discord servers, LinkedIn threads, support groups, the comment sections of the tools people already use. This is where you learn the words customers actually use for the problem, and the workarounds they are proud of. You will need both for the conversations that come next.

§Customers last. Not because they matter least. Because by the time you reach them you know the market, you know the naive questions, and you speak the language. The interview becomes a test of what you think rather than a tour of what you do not know.

§ 1.2Before you look: the three planning questions#

§Research without a plan produces a folder. A plan is three questions, written before anyone opens a browser.

§What do we assume? Not the vision, the assumptions underneath it. "Independent restaurant owners lose sales to stock-outs." "Developers would pay for an assistant that knows their whole codebase." "Small law firms cannot afford the research tools large firms use." Write them as claims that could be false.

§Who do we think has the problem? A description specific enough that you could tell whether a given person is in or out. "Small business owners" is not a target. "Owners of two to five restaurants who still count inventory by hand at closing" is.

§What would change our mind? This is the question teams skip, and it is the one that makes the rest honest. If you cannot say what finding would make you drop the idea, you are not researching. You are collecting reasons.

§ 1.3The checkbox trap#

§The failure this chapter exists to prevent is the team that does all four sources and learns nothing, because each was done to be able to say it was done. Ten competitor sites skimmed. Two experts pitched to. A subreddit read for an hour. Eight interviews in which the founder talked for thirty of the forty-five minutes. Every box checked, every assumption intact.

§The defense is the third planning question. If you wrote down what would change your mind, you can check at the end whether you looked for it.

§ 1.4What you leave with#

§A one-page research plan. Assumptions as claims that could be false. A target described tightly enough that you can say who is out. One line per source saying what it is for. And a sentence, in plain words, describing the finding that would make you stop. Pin it where you will see it. Chapter 7 will ask you to read it again.

1. A hierarchy of sources