Chapter 3: Asking Important Questions

Core Idea

Unbiased questions are useless if they’re trivial. In every conversation, ask at least one scary question whose answer could destroy your current plan. Start broad before you zoom in, and check both customer risk and product risk.

Frameworks Introduced

The important-question test: a question is important if its answer could completely change or disprove your business. If you get an unexpected answer and it doesn’t change what you do, the question wasn’t important.
Failure and success thought experiments: imagine the company has failed and ask why. Imagine it as a huge success and ask what had to be true. Then find ways to learn about those critical pieces.
The delegation trick: when you keep avoiding a question, imagine what you would tell someone else to do if you delegated it, then do that. You’re allowed to ask about money.
Love bad news:
- When to use: when a response is lukewarm, such as “Umm, I’m not so sure about that” or “That’s pretty neat.”
- How: don’t “up your game” and pitch until they say something nice. Ask one or two questions about why they don’t care: Is the problem not a big deal? Are they different from your ideal customers? Do they dislike this particular implementation? Are they tired of pitches? Are they just tired today? Then thank them and leave.
- Why: if you have $50k, spending $5k to learn a path is a dead end is a great result. You keep the rest and know more.
Look before you zoom (avoid the premature zoom):
- When to use: before asking detailed questions about a specific problem.
- How: first check whether they care about the wider area. Open broad (“What are your big goals and focuses right now?” or “What are your big problems right now?”). Zoom in only when they raise the topic without prompting, or when you already know it’s a top-three must-solve problem (for example, marketing for small businesses).
- Failure mode: a premature zoom produces answers that look like validation but aren’t. People will always name a “biggest problem” when asked, even in an area they don’t care about.
“Does-this-problem-matter” questions (painkiller or vitamin?): “How seriously do you take [it]?” “Do you make money from it?” “Have you tried making more money from it?” “How much time do you spend on it each week?” “Do you have any major aspirations for [it]?” “Which tools and services do you use for it?” “What are you already doing to improve this?” “What are the 3 big things you’re trying to fix or improve right now?” These check whether they spend money, make money, rank it in their top three, and are actively looking for solutions.
Gaze upon the elephant: list every failure point and don’t focus only on the most interesting one. Example: poor schools’ teachers are overloaded (true), but the schools may not have budget to pay you.
Product risk vs customer/market risk:
- Product risk: Can I build it? Can I grow it?
- Customer/market risk: Do they want it? Will they pay me? Are there lots of them?
- “If you could get me more X, I’d pay you” moves the risk onto your product. It confirms something that was never in doubt. With heavy product risk, conversations can prove less, so start building earlier and with less certainty. Still talk to people, so you don’t succeed at building only to find nobody will pay.
Prepare your list of 3: plan the three most important things you want to learn from each type of person (customers, investors, industry experts, key hires). Decide them calmly with your team, update them as answers come in, and don’t repeat questions you already have solid data on. The list also lets you ask your key question straight away in a chance encounter.

Key Concepts

Premature zoom: asking detailed questions about a problem before learning whether the person cares about the area.
Lukewarm signal: polite indifference; reliable evidence that they don’t care.
Painkiller vs vitamin: a must-solve-now problem vs a nice-to-have.
Elephant in the room: the critical risk you avoid asking about.
Product risk: the risk lies in whether you can build and grow the product.
Customer/market risk: the risk lies in whether people want it, will pay, and exist in numbers.

Mental Models

Think of trivial safe questions (like asking someone’s age) as spinning your wheels. They don’t bias, but they don’t move you forward.
For learning (not selling), “meh” is a great response because it is reliable.
Use the advertiser test for product risk: advertisers will always pay for page views if you have them. You don’t need conversations to confirm it; the risk is in getting traffic. Video games are almost pure product risk.
The author never asked to speak to a key customer’s lawyers about a legal grey area he was nervous about. Avoiding that one question cost at least half a million dollars.

Anti-patterns

Asking only comfortable questions: unplanned conversations drift toward trivia.
Pitching to overcome a lukewarm response: unless they’re holding a cheque, all you gain is a false positive.
Ranking exercises on unvalidated problems: “Rank convenience, personalisation, novelty, cost” assumes they care at all.
Taking “If you can build it, I’ll buy it” as validation: the farm fertility tracker founder spent three months on such conversations, then couldn’t build the product. Bar-promotion apps hear the same from bar owners but need a huge consumer audience first.

Rules of Thumb (author’s wording)

You should be terrified of at least one of the questions you’re asking in every conversation.
There’s more reliable information in a “meh” than a “Wow!” You can’t build a business on a lukewarm response.
Start broad and don’t zoom in until you’ve found a strong signal, both with your whole business and with every conversation.
You always need a list of your 3 big questions.

Worked Example

A fitness app conversation, zoomed too fast and then fixed.

Bad version: “How often do you go to the gym?” “Not really ever.” “What’s your biggest problem with going to the gym?” “I guess the time it takes to get there.” “Rank convenience, personalisation, novelty, and cost.” He ranks them, hypothetically. You pitch your home-workout app. “Cool, I’d love to try it when it launches.” You send a beta key that he never uses. It looked like it went well, which is why it’s so harmful.

Fix 1, dig into the “no”: “Why not?” “It’s just not something I’m that worried about.” “When’s the last time you did try? A gym, jogging?” “Sports in high school. Chasing the kids gives me enough cardio now.” Thank him and leave. You learned what you came for.

Fix 2, start broad: “What are your big goals and focuses right now?” A promotion, a new house to fix up, more time with his spouse. “Is getting healthier on that list?” “I’m feeling pretty good at the moment.” Not a customer. A leading question was acceptable here only because you were about to end the conversation anyway.

The speaker marketplace (an elephant example). A speaker earns $2–3k per talk and has a mediocre agent, but most gigs come through his blog and past talks. You pitch a marketplace that cuts out agents. He says that if you could get him more or better-paid gigs, he’d drop the agent, pay you 20%, and bring friends. It sounds like a commitment, but it only says he’d like more money. The real risk is whether you can attract enough paying event organisers, which is product risk he can’t answer.

Key Takeaways

Before each batch of conversations, write down the questions that scare you, and ask them.
Treat lukewarm responses as solid data. Ask why, then move on.
Start broad, and zoom in only after they raise the topic themselves.
Use the does-this-problem-matter questions to tell a painkiller from a vitamin.
Name all of your failure points, including product risk, and say which ones conversations can’t test.
Keep a list of 3 big questions for each type of person, and update it as you learn.

Connects To

Ch 1: The Mom Test makes questions unbiased; this chapter makes them important.
Ch 4: the list of 3 lets you skip formal meetings and just ask.
Ch 8: the list of 3 is the core of prep and review.
Pre-mortem (Gary Klein): the failure thought experiment works the same way.