On the advice of Fred Wilson from AVC I picked up two new books this week:

Customer Development 101 – Talk to Humans

Talking to Humans is a quick read – a couple hours or less – and caught my attention with the simple title and summary. I’m a human. I’m a fan of talking to other humans. Sign me up!

It’s popular in Lean Startup conversations to preach customer development and “getting out of the building.” You need to talk to real (potential) customers before you build. Less popular a topic, though, is giving the entrepreneur advice on the right and wrong way to talk to customers.

Looking at Less Meeting, for example, we’ve always scored highly when it comes to quantity of customer conversations. However, after reading Talking to Humans I see mistakes we’re making and how the quality of the conversation is just as important as the quantity of conversations.

So how should we start talking to humans?

1. Focus on Past Experiences, Not Future Speculation

People love to speculate about all the features they would want if they could have them. Which is great. Except that it’s not great at all.

Take Less Meeting Teams for example. This is a page where you can share notes and action items across your teams’ meetings, even meetings you’re not a part of. Users rarely visit this page though. And looking back at how we asked customers about Teams we’re guilty of leading the witness. I say this because we tell them about this concept of Teams and ask them how they would like it to be better. That’s such a bad approach, most notably because of how bad customers are at guessing how they’d want to use Teams.

The reason is that speculation is far removed from reality. Instead, a much more accurate predictor of reality is past experience. History is a better indicator of future behavior than anything else, so we should be asking customers about the problems they’ve experienced before seeing meeting information across their teams, and how they’ve tried to fix this problem in the past.

2. Interview Many Customer Segments

In Talking to Humans Giff uses an example of a fake pillow startup company to convey his points. During their customer discovery they go to a Bed, Bath & Beyond. And then talk to people at a large department store. They talk to their friends. They talk to their friends’ friends. They do online research. They talk to people at a coffee shop. They interview a bunch of recent grads. Then interview people 5-10 years out of school.

They talk to a lot of different customer segments.

At Less Meeting we talk to a lot of people too, but nearly all of them are existing customers. 

When you’re doing customer development you need to get creative, get scrappy and get some hustle. You won’t reach statistical significance in your observations – and that’s ok. However, you still want to find patterns to increase your chance of success. To do that, you need to observe people in different environments & in different backgrounds.

Bonus: If you’re a B2B company remember to talk to the many customer segments within a single organization – the strategic buyer (usually the champion), the economic buyer (usually the check writer), the technical buyer (usually an IT admin), etc.

3. Use Customer Interviews to Validate Assumptions

Here’s an important detail. In order to validate assumptions in your customer interviews, you need to actually know your assumptions.

In Giff’s story of the fictional pillow startup, the team completes a Business Model Canvas. Don’t know what that is? Click that link now. I’m watching you – go click it. Do it!


Here’s the million dollar lesson: by creating a canvas for your product (and customers!) you’ll inherently uncover the key assumptions. Assumptions lead to risks. Risks turn into issues. Issues that weren’t accounted for will cause your startup to fail. It’s that simple – failure to prepare for assumptions in your business model canvas will bite you in the ass.

The target customer segment for Giff’s pillow startup was recent college grads. This is a big assumption, so the team talked to a bunch of recent college grads. When they did, they found out that people a year or so out of school don’t actually have much money. And without much disposable income they prefer to keep their old ratty pillows they already had.

But by also interviewing other customer segments (such as people 5-10 years out of college) they found that there is a customer segment that IS willing to spend money on a high end pillow. And what’s more, by focusing on past pillow buying experiences instead of future speculation they learned WHEN people typically buy pillows. (Hint – the answer is during big life events like getting married or buying a house.)

Don’t be lazy…get with your co-founders and create a Business Model Canvas. Then use the canvas to uncover your lurking assumptions. Next, get out of your frickin’ office and start talking to real humans. And when you do, make sure to observe and interview them the right way.