#123 - Harry Zhang and Kevin Hale

**Craig Cannon** (0:00)
Hey, how's it going? This is Craig Cannon, and you're listening to Y Combinator's podcast. Today's episode is with Harry Zhang and Kevin Hale. Kevin is a partner at YC. Harry is the co-founder of Lob. Lob makes it possible for enterprises to programmatically send physical mail. They were in the summer 2013 batch of YC. You can try Lob at lob.com. Harry is on Twitter at Harry Zhang, and Kevin is at I Like Vests.
All right, here we go. Today we have Harry Zhang, co-founder of Lob. And Lob makes APIs for companies to send letters and postcards.
So, Kevin, I have a question for you.

**Kevin Hale** (0:39)
I'm trying to think back to when you guys applied to YC. You didn't have almost anything.
Like, I would say it was fairly embarrassing.

**Harry Zhang** (0:50)
I guess you could say that. I mean, honestly, we were scrambling. Weeks before the interview, they tried to come up with something that we thought was worthwhile for demoing. But I think the important thing when you're starting a company, and for us, it's thinking about the problem that you're working on versus having a bunch of stuff.
And I think for us, we had done the due diligence beforehand. We talked to a lot of customers. We had a good sense for what the core problems are, but we didn't have the exact product just yet.

**Kevin Hale** (1:17)
What was that MVP when you guys applied to YC at the time?

**Harry Zhang** (1:21)
I think back, and it's pretty funny just thinking about this. We literally had an API that took whatever you put into it and spit it right out to you.
And that was our MVP, and we were secretly hoping.

**Kevin Hale** (1:31)
It just mirrored the response back.

**Harry Zhang** (1:32)
Yeah, it just mirrored the response back, and we were just sort of hoping that we wouldn't have to demo, because honestly, we hadn't built a lot of product yet. We didn't know what we wanted to build, and we've always operated from the penalties. We want customers to pay us for something before we actually go and do it.

**Kevin Hale** (1:47)
What's interesting to me, as someone who's focusing on design interfaces at YC, is that I kind of really love API companies because I think a lot of people think of it as like, oh, it has no sort of interface. But to me, the interface is just documentation. It's like very pure.
And so I'm always interested in, when I look at an API company, is looking at, oh, what are they doing here to say, this is how we want to build community around this? Because what they're trying to build out is that this is something that people who build stuff want to use. And so what are the ways they're going to communicate that this will allow them to do something sort of cool or solve a problem, et cetera?
And I feel like you learn a lot about people's thought processes and also how good someone is as an engineer or programmer by how thoughtful that documentation is. I think Stripe is a classic example of this. But for you guys, I remember, you had the basics down, but it was clear that you guys knew enough to be considered to anyone that would want to be using it, even if it was their first time ever writing any API call.

**Harry Zhang** (2:55)
Yeah, yeah, that's definitely right. I mean, I think documentation to this day is still like the number one thing. And I think we take a point in pride in making sure that our documentation is something that every single developer's gonna use.
And I see it as friction. Like when your core product is an API, you better have like the best documentation on the planet.

**Craig Cannon** (3:13)
And so what was the insight you guys had early on that made you convince that you wanted to follow this path?

**Harry Zhang** (3:18)
Yeah, that's a good question. So it actually started from like the problem and it started from my time at Microsoft. So, admittedly, I did not think I was gonna end up in a business that's focused on mail, nor did I think that at a job at Microsoft, at a company that's 100% tech-based, one of the main things I was gonna be working on was actually a direct mail campaign. So that's sort of like how we initially got started with the problem. And it actually started from a place of complaining.
I was driving back from a ski trip with my co-founder, Lior, and I was complaining to him about this project that I was working on at Microsoft. And essentially what we were doing is we had come up with a system for sending out these like very custom training materials and an invite to like a webinar online. But for a number of different reasons, we weren't allowed to do that using email and telesales was too expensive. So the thing I was doing was actually building a direct mail program. That was one of the most effective things that we've done to date. And honestly, it was a little out of, we didn't have any other options, but when it worked really well, I got a huge budget and was asked to sort of scale across a much bigger set of customers and different use cases and more products.