The Mom Test notes

27 Oct, 2025

My notes from the book. A later write up could be as a graphic or process diagram I'd ideally follow, maybe as an interactive graphic.

One way I think about all the processes this book proposes: it can all contribute to creating an information machine which is constantly updating beliefs on how likely each idea is to succeed, where to invest effort next (inc when to put effort into getting more info vs buildings vs other things). I don't mean information machine as a computer here, but a system I and my company use to spend our time approximately optimally. I'm not sure I'd follow such a system properly alone, but as a team we might stick to the process well, checking in daily and progressing stuff, keeping each other honest and correcting errors.

Much of the book is specific tips for your conversations. I'd love to internalise all these. Unfortunately, reading the book hasn't fully internalised them for me (suprise suprise!) I think I have it in me to do this stuff really well, however I would learn much faster with a guide, mentor or peer to peer feedback.

The very high level idea:

• Talk about their life instead of your idea
• Ask about specifics in past instead of generics or opinions about the future
• Listen plenty

Customer learning can move really quickly when you’re doing it right

One key idea is to make it all about them. We all know most people are people pleasers, however getting feedback on your product sets this off to another level. The book posits that if people get even a tiny sense that you might be emotionally invested, then they will stop telling the truth and start saying nice things. This can be very dangerous: it gives you the wrong impression that they might be a customer, but they almost certainly won't.

Specific tips I noted as I read:

• Ask for specific examples in the past to get real concrete data
• Attack generic answers by asking for specific examples
• A conversation is useful if it gives concrete facts of customers lives and world views
• Generic ideas always seem good: need to make them specific 
• At some point you need to mention your idea, but generally one does this too soon rather than too late
• ‘What would your dream product do’ is an ok question, but only if you ask follow ups (it sets these up well)
• Ask for the implications of a problem someone describes - they might not care enough to do anything about it
• Ask what else they’ve tried to solve their problem. If they haven’t looked for anything yet, they probably don’t care enough to buy your thing
• Ask how they are dealing with the problem now. Inc any expenditure so can get ballpark of what they’d pay
• At end of convo: ask who else I should talk to and anything else I should ask

Gives 3 kinds of bad data:

• Compliments. "It looks great!"
• Fluff (generics, hypotheticals, the future). E.g: "Of course we'd want to use it one day"
• Ideas. Says their role is to talk about their problems, then your role is to have solutions. I think it's sometimes fine if they have ideas for us to explore them. The book suggests that people might want a feature or suggest an idea for a strange reason: better to understand the problem they're trying to solve with the idea, and then there might be a better way to achieve that. Asking why helps here.

Suggests getting any of the 3 kinds of bad data should be a sign to change the direction of the conversation.

Ignoring compliments should be easy but due to human nature it often isn’t so.

• Once you have a specific case of something interesting to discuss (a problem, a workflow) can help to ask them to walk you through it.
• When using generics people tend to describe themselves as they want to be, not as they truly are. Need to be specific to bring out the edge cases where we can add value.
• If someone is being flaky can put a decision to them - if they don’t care enough to try solving their problem today that won’t change tomorrow
• Understand the motivation behind requests
• If you mention your idea, people will try to protect your feelings
• Posits you should be terrified of at least 1 question you ask in each conversation
• Lukewarm responses are useful: they tell you the person doesn’t care
• Ask for major annoyances, costs, joys, goals, focuses
• If you have more product risk (ie it’s unclear if it’s possible to make your product work) then start testing that earlier in the process than you would if all the risk was market-based
• The most valuable resource of a startup is the founders time: you have to put yourself where you matter most
• Many people become more formal to deal with ambiguous or awkward situations (uses example of asking someone out). Avoid this.
• The best user chats feel like a pleasure for both parties
• In person chats are best for making stronger connections

One sales and meetings:

• When you fail to push for advancement (next stage in sales funnel) you get zombie leads who keep saying nice things but never go anywhere
• Says meetings don’t ‘go well’ , they succeed for fail. If you advance it succeeds, if they reject that’s ok too, if you don’t even ask that’s a failure
• Them committing something (time, money, reputation) are all good signs
• The more someone is giving up the more seriously you can take their validation
• It’s not a real lead until you’ve given them a concrete chance to reject you
• Says you should decide in advance what 3 things you want to know from a group (eg customer, investor, potential employee)
• Learning about a customer and their problems works better as a quick and casual chat than a long formal meeting (says you could have 12 such chats at a meet-up event)
• Being too formal is a crutch we use t

Your first customers:

• In finding people to chat to, says be open to serendipity everywhere (old friends, parties... anywhere outside the context you'd expect to find people)
• First customers, to buy from a start up, have to really really like what you’re doing (in b2b more so I assume). Look out for people who are particularly emotional about what I’m doing. They are the ones who commit before it makes rational sense to do so
• In early Stage sales the goal is learning, revenue is a side effect
• The goal of cold conversations is to stop having them (because you get well known)
• Says worth choosing customers you admire and enjoy being around
• If a Q can be answered with desk research, do not ask as part of a conversation. And for prep says who start up team should be there (good for collective knowledge imo, and good for improving process itself too)

Things you might do to get leads in industry:

• Organise a meet up for people in that field. Says it’s way better than just going to another meet up and you get way more from it. Also says nobody does it and they don't know why. I would consider it.
• Teach (eg free teaching at conference or online , meet up, etc)
• Set up other networking events (eg a ‘knowledge exchange between universities, Doug’s ai breakfasts)
• Get industry advisors 
• Academics you know (or ones u don’t know as their emails are always online)
• Investors 

More tips on finding leads:

• Says because of 6 degrees of separation, you can find anyone you need if you ask around a few times
• In colder calls and initial contact don’t be looking for customers. Be looking for more advisors and knowledgeable people excited for the area
• Get specific about your ideal customers - do customer segmentation

Pages with useful references:

• Highlighted p32 questions to dig into feature requests and emotional signals
• Highlighted p49 for ‘does this problem matter’ questions
• Highlighted p87 of steps to frame a meeting
• Highlighted p100 for Qs to ask self to help narrow down to ideal customer group
• P110 for notation for capturing important stuff…. Says taking good notes can make it harder to lie to yourself. Highlighted p115 for more things to do before during after convo 
• P120 for cheat sheet (list of example Qs)