In this episode of the My First Million podcast, the host and Sam Parr discuss the critical role of deep customer discovery in early-stage startups. They begin with Paul Graham’s essay “Do Things That Don’t Scale,” recounting how Airbnb’s founders, prompted by Graham, went to New York and personally took professional photos of listings, a tactic that dramatically improved bookings and later scaled into a photographer fleet. Parr shares his own experience of extracting high‑value insights by talking to just ten users, illustrating that a tiny sample can reveal patterns for thousands. The conversation then shifts to Twitch’s early days, where CEO Emmett (formerly of Justin.tv) kept meticulous notes from streamer interviews. He used a simple three‑question framework—what users like, dislike, and what would make them switch—to iterate product features, and he found that after about six conversations the same themes repeated, allowing him to predict future feedback. The host highlights Emmett’s aggressive “bulldozer in a field of flowers” mindset and the importance of sharing early‑stage resources like pitch decks. Finally, both agree that the best practical guide for conducting effective customer interviews is Rob Fitzpatrick’s short book “The Mom Test.” The episode emphasizes actionable, low‑scale tactics for learning from customers that can later be systematized.
Chapters
You know, there's probably like 10 blog posts that matter the most if you're going to do a startup. One of them is the Paul Graham essay, uh, Do Things That Don't Scale, uh, where he's talking about with Airbnb, uh, early on. And if you're kind of in the startup world, just fast forward 30 seconds, but if you're not, you'll like the story, um, if you haven't heard it. So Airbnb is in Y Combinator, they're They're struggling. Airbnb is not really working. They have like, I don't know, 20 customers or something low. And they go into Paul Graham's office. He's the advisor from Y Combinator. And he's like, okay, so tell me how it's going. And they're like, well, you know, it's small and not really growing. And he's like, okay, but tell me, what have you learned from your customers? You've been out there. What'd you learn? And they're like, well, you know, most of our customers are in New York and they're putting their apartments up. He's like, they're in New York? They're like, yeah. Like most of your customers are in New York? He's like, yeah, he's like, what are you doing here? Like, go to New York, you need to go to New York now. And like, why are you in San Francisco if your customers are in New York? So he sent them over there, and what they did, which is something that nobody else was doing, was if you ever saw an apartment listing before, the photos looked like shit because it would be somebody on their crappy mobile phone at that time taking a photo of a dirty apartment and listing it for rent. And so it was like the worst sales pitch. And so they started themselves going in, they said, hey, can we send up, they would tell their customers, Hey, we want to do professional photos of your place. It increases bookings by 40%. Are you open to that? And then they would show up as the professional photographer and be like, hey, by the way, I'm the founder. And they would take photos and the Airbnb listing started to look great. And everyone told them this won't scale and they also thought this won't scale and to an extent it did because even now today there's a sort of fleet of photographers that will come and take photos of your place to list it because it was just that important and they found a way to make it scale. Scale.
I think that's a great story. Another thing that shocks people is I think even if you have hundreds of thousands of users, maybe as high as millions, you could really— I, I've done this with us. I'll just call and talk. I only need to talk to about 10 people.
Right.
And I can see the pattern that represents tens of thousands or hundreds of thousands of people.
Right.
I could— you could just tell you. So like what you could do is A/B test shit and wait. If you're just starting, you're not gonna— it's not gonna be significant. And that could work fine. If you just call 10 people, you really get the information you need.
Totally. I was talking to Emmett. So Emmett's the CEO of Twitch where I'm working. And, um, one, you know, being at a big company has all these different weird things that aren't really related to my core interest, which is startups. So I ask Emmett all the time, just tell me about the early days. And so I asked him, I said, hey, uh, when you guys were pivoting from Justin.tv to Twitch, was that obvious? He was like, no, dude, at Justin.tv, like less than 2% of the watch time was on video games. It wasn't obvious. That this was the thing to do. He's like, I just liked it and I thought there was an option. And he goes, and then I talked to a bunch of customers and I said, do you have your notes from that? And he's like, yeah, I do. And so he sent me his notes from the initial conversations he had with all these streamers, all these customers.
Can we publish those?
No, no, I don't— probably not. But I, I was reading through them and it was fascinating. So first was just like, first lesson was just go ask the founder or CEO about the early days. I bet you they have a bunch of resources that are awesome. Like he sent me the old pitch decks and stuff like that. I think I sent you one slide, which was my favorite slide. The title of the slide, because it was a progress update for their early investors. Yeah, and it just said, we're like a bulldozer in a field full of flowers. And he just— and then it was a list of all the shit they got done that month. And I was like, this attitude, like, if I said that today at Twitch, it would be like, what is this maniac saying? Like, why is your slide so aggressive? But that's how they were early on. So then he had all these notes of his customer conversations, and he always tells everybody at Twitch, he's like, Go talk to people. By the 6th conversation, you'll hear the same thing over and over again. It takes 6 phone calls basically to figure out the pattern. And by the, by the 6th one, you'll know. And, uh, you'll, you won't even wanna do 50 conversations. You'll be like, all right, I, I can predict what this person's gonna say before they even say it after 6.
That's awesome. I think we just got ourselves another clip.
Clip it.
Uh, what do you wanna, wanna, uh, first of all, share that deck with me again and also ask a ask him if we can make that public.
Yeah, I will ask him actually, cuz it's pretty cool.
What do you want to get— what do you want to do?
Uh, the one thing I can share out of that, by the way, he asked 3 ques— he's like, I said, you know, oh, customer interviews, is there a whole skill set to learn? He goes, no dude, I asked 3 questions. I said, what do you like about your current platform? What do you dislike about your current platform? And what will it take to get you to switch to Twitch? And that's what he asked every single customer. And then he just heard the common things and he would build it and he'd go back to him. He said, You told me this is what it would take for you to switch. 2 weeks later, here it is. Are you ready? And then sometimes they're like, no, still. He's like, okay, cool. What's next? And he would just like keep doing that. 3 questions.
That's badass. Just to wrap that up, if you wanna learn about this stuff, I think the best book I've ever read on this is called The Mom Test.
Yeah. I love that. Jack recommended me that book.
Yep.
It's fantastic.
It's the best thing.
Tiny book too.
Like, that's a long blog post.
50 pages or something.
That book, um, in terms of your business life, it might change Yeah, the mom test.
