I recently watched this talk by Eric Migicovsky at Y-Combinator on how to talk to users.

Here are my notes.

👾 Initial thoughts from Sean

• Talking to users is an interesting topic because it seems like something that should happen at a later stage. I worked at a startup where we had regular discussions with our most active users about our SAAS platform, the features we were building, and what was and wasn’t working well for them. To me it seemed like something that happens later on in the startup’s lifecycle when you…y’know, have users.
• Thus, I almost didn’t watch this video because I haven’t even incorporated my startup yet, let alone built the platform. But “Talk to users” is a bit of a misnomer because it actually includes talking to people who are not yet your users. It’s critically important to talk to prospective users and even just advisors in the industry you are targeting when you are still evaluating whether your idea is viable. This is how you prove out the concept before you’ve built anything.

📕 Now, the notes:

• It is the job of the CEO to talk to customers. This cannot be passed down the chain.
• If the engineers don’t have the skills to talk to users, they need to also develop those skills. There should’t be a go between.
• Y-C’s core teaching is that you mainly need to do two things: build your product and talk to users.
• A lot of the info from this talk is also synthesized nicely in the book “The Mom Test”. The general idea is that you need to be able to explain your idea to someone’s mom.

😳 3 common errors founders commit when talking to users

1. We talk about our idea

• During a user interview the goal is to listen, not to talk about your idea.
• When you talk about your idea you are moving the conversation into a social discussion where they will try to support you rather than give you useful feedback.

2. We talk about hypotheticals

• We ask questions like “if we built this feature, would you use it or pay for it.”
• It’s better to talk about specifics in their life.
• Recently, when exploring an idea I asked “In the past year has your company done X?” This adds a relevant time restraint, showing that the question is one of fact and not one of opinion.

3. We talk more than we listen

• This…is a problem in every aspect of human existence, not just among startup founders. But that’s a tangent I’ll stay away from for now.
• In a user interview, try to restrain your need to talk.
• You might unintentionally steer the conversation in a certain direction and miss out on important insights you would get from your users.
• If you steer the conversation, you will only learn about known unknowns, but the most valuable thing when exploring startup ideas or getting feedback on your platform are the unknown unknowns, the things that you don’t know your users are thinking about or worrying about.

✅ Five great (open ended) questions everyone can ask:

1. What’s the hardest part about the problem you are trying to solve?
2. Tell me about the last time you encountered this problem.
3. Why was this hard?
4. What, if anything, have you done to try to solve this problem?
5. What don’t you love about the solutions you’ve already tried?

Eric fleshed these questions out with the example of dropbox and working on a group project in college. I didn’t take any notes on this because it seemed fairly obvious and, if I’m being honest, I don’t think dropbox is a very useful example at this point. Cloud data storage is too ubiquitous to put you in the right mindset. It would be better if we asked these questions around a startup idea that failed or pivoted. Or at least one that is more recent (e.g. SpotHero).

⏰ When in your startup’s lifecycle to interview users

Next Eric outlined the three stages where user interviews will be most useful: the idea stage, the prototype stage, and the launched stage.

For me, most of my experience working at startups has been talking to users during the launched stage. Now that I’m trying to prove out my own startup ideas I’m mainly talking to (potential) users in the idea stage.

🗣️ Idea Stage

• Find first users with problem
• Friends, coworkers, intros
• Drop by in person
• Industry events

idea stage tips:

• take detailed notes
  • You won’t know what key facts are important until later.
• Keep it friendly and casual
• Be cognizant of the other person’s time. This is one of those times where you need to restrain your desire to talk or pitch your idea.

🚗 Prototype Stage

• identify best first customers
  • At this stage, you can use user interviews as a way to create the ideal customer profile that you will use to find your best first customers.
• Find numerical answers to:
  • How much does this problem cost them today?
  • How frequent is the problem?
  • How large is their budget?
  Here I would caution you to resist the urge to just ask them these questions outright. These are more things that you should interpret from what you learn from them. On a case-by-case basis you may be able to ask some of these outright, but do so with extreme caution, you might come across as rude.

🚀 Launched Stage

• Iterate towards product market fit.
• PMF: When you have made something people want. When you don’t have to push the product on customers.
• “how would you feel if you could no longer use X?”
  • Three options: Very disappointed, somewhat disappointed, not disappointed.
  • Measure the precentage who answer “very disappointed”
  • This shows if you’re trending towards product market fit.
  • if 40% or more says very disappointed, you’re on track to product market fit.
• To do this you need to already have users.
• Ask your users for their phone number during signup. That way if they are encountering a weird problem you can just pick up the phone and call them.
• Don’t design by committee. Find a way to get users to opt in for a new feature even if you haven’t built it yet. that will help tell you if it’s viable.
• Discard bad data. Bad data mostly means compliments. Talking to users is not about making you feel good.
  • If you want to feel good, go home and have a beer and an edible. This is about improving your product and the only way to do that is to set your ego aside and get real information.
  • The other type of bad data is fluff. Don’t talk about what the product may be like in the future. Try to stay in real tangible data points.

❌ Bad questions

Lastly, I would also like to add some bad questions you may be tempted to ask. This is entirely from me and is not something that Eric dug into so much.

1. How would you rate our product, idea, etc on a scale from 1-10?

• This seems like a fine question but for the user it comes across as slightly confrontational. They will be reluctant to give you a 2 even if you deserve a 2.
• At the same time, you would know in your mind that you want to hear that your idea is a 10 out of 10. Having something that you want to hear will make it harder for you to accept and interpret anything less in a constructive way.

2. What features are missing?

• It’s hard for people to imagine what’s missing.
• At best here you’ll get no answer, at worst you’ll get a rabbit-hole that doesn’t actually translate into anything useful for 99% of your customers.

3. Would you pay more if…?

• There are ways to have constructive conversations about money and value but you have to step very lightly.
• You don’t want to give the impression that they are a meal ticket for you.
• You are interviewing them because you value their advice and opinion and not because you are trying to sell them more features.

About me

Hey, I’m Sean. I’m an aspiring startup founder and I decided to #LearnInPublic by sharing my notes as I learn.

My background is in data science, astrophysics, and growth hacking with some entrepreneurship scattered throughout my timeline.

If you found this helpful feel free to add me on linkedin or twitter (<at>Seanlikesdata) and let me know.