Parker Conrad's hard-won billion dollar startup lessons: Do unscalable things, but then actually scale it (Part 1)

New Vlog this morning! It was a real treat to sit down with Parker. 

Transcript below

Today we're going to hang out with Parker Conrad. He created Rippling. He's one of the best product-focused CEO's I've ever met. Today he's going to share some of the most hard-won startup lessons that I've ever heard. And he hasn't talked about it anywhere else.

Let's go check it out. This is part one.

Garry: Thank you so much for hanging out.

Parker: Yeah, no, thanks for coming by.

Garry: So Parker, what is Rippling?

Parker: If you're a founder or CEO, one of the things you probably noticed at your company is that there's a lot of just irreducible admin work involved in running a business. Could be getting hired, it could be getting a raise, a promotion, getting married, having a kid, moving to a new address, eventually leaving the company, and each of those life cycle events for an employee create, have this set of implications across all or some subset of your different business systems that today, you need to handle manually and by hand, but with Rippling, we automate that end-to-end.

So, we get people added to all the right email lists, the right Slack channels, we get them a computer, all the right softwares installed, they're enrolled in the right insurance benefits, get 'em paid correctly.

Garry: Taking a step back, why did you start working on this to begin with? Why startups? Walk me through your path to here.

Parker: Actually, in 2006, I was working in L.A. for a pharmaceutical company, and I had a great life. I was living two blocks from the beach with my beautiful girlfriend, now wife, had this great commute up and down PCH, was getting lots of atta-boys at work, but I sort of sat around, I looked at it, and at the company I was working at, you had all these different levels that you needed to move up in your career, and you had to spend two years here and two years there.

At a certain point, I realized that, if you added up all the time it would take to get anywhere in your career, I was like, I'm gonna be dead before I accomplish anything. And around the same time, my roommate from college wanted to start a company, and so we did the absolute worst thing that you can do to start a startup, which is, we said, we decided we wanted to start a company, and then we sort of started looking around and thinking about well, what should we do?

We came up with a bunch of stuff that it turns out nobody really wanted. We built a wiki for stock research.

We ended up pivoting the company every six months for seven years, we were constantly trying to sort of find something.

Garry: I remember, because my first interaction with you was actually maybe renting space from you down the street in South of Market.

Parker: Oh, is that right?

Garry: At the office that's now a restaurant? There's this tiny little hole in the corner of the basement that could've been our startup.

Parker: So yes, I was at that company for seven years of slow, grinding failure.

We talked to 70 different investors. Everyone turned us down. I mean, we raised a little bit of money, and we made it last, and we stretched it out, and we were constantly a month or two away from missing payroll.

And eventually, my co-founder and I had a falling-out, and I left the company, and I started Zenefits the day that I incorporated it, basically my last day at SigFig.

Garry: Yeah, I remember, and then you applied to YC, and I remember very distinctly reading your application. The thing that really jumped out at me was that you had identified that, basically, regulation was coming to sort of shake up this ossified industry of insurance.

Parker: So, when I started Zenefits, no one was looking at the health insurance market. Yeah, I talked to a few investors. But this wasn't one of the trends that people were thinking about, but it seemed like an industry that ran on information.

There was no physical goods exchange. When you think about health insurance, you gotta enroll, they have to make payments to doctors, but it seemed like it's the perfect industry to be run by technology.

Garry: Totally. This is a very common thing, actually. People always ask me, as an investor, what's hot, what's the trend, and my reply is always, don't ask me. I can only tell you what happened six months ago or three months ago. You have to go and find something that's happening that nobody's working on.

Parker: That's right. 

Garry: What is it that really speaks to you, that you think is a real opportunity? Only then can you find something that could be really big.

Parker: So, the reason I started Zenefits is that I had dealt with setting up insurance for my previous company, and it was super painful. The only way to enroll people was via the fax machine, so every time we hired a new employee, I needed to stop off at Kinko's to send in a fax, because we didn't have a fax machine, and that was the only way to get someone enrolled. And so, I was like, gosh, there's gotta be a better way to do this online.

Being a broker, being the person that helps companies get up and running with insurance, and services, the plans, was just this incredibly lucrative revenue stream. I mean, I have been beaten up so much trying to raise money that I had sort of given up on being able to do that, and so I figured if I can get into YC, maybe I'll be able to raise a few hundred thousand dollars.

There's so much cash in being a broker that maybe I just never raise again. Maybe it'll be a really big company, but honestly, I was thinking, man, maybe I could just be a reasonably successful tech-enabled insurance broker and make it work as a business.

Garry: Yeah, totally. But it turned out that you were able to grow faster than I think anyone in Silicon Valley--

Parker: Yeah, it turns out that's the kind of business a lot of companies wanna invest in.

Garry: A lot of investors wanna invest in things that are growing massively.

Parker: When I started a company and the underlying metrics were no good, there's nothing that we can do to raise.

Garry: In order to support that growth, there's sort of two pieces to the full-stack startup.  It's the software that automates, and then there's the people.

Parker: Fundamentally, my view of what went wrong with Zenefits is that I made a decision very early on that we were gonna scale the business faster than our engineering team could really support it. And so, we thought, look, what we're gonna do is not everything's in place yet, not everything's fully automated. 

What we're gonna do is, companies are doing all this admin work; we're just gonna do it for them.

And it was a classic YC do things that don't scale, which is absolutely the right way to start a company.

The corollary to that, though, is eventually, you have to scale them.

Garry: That's the part two. And the majority of companies actually don't get to that point, so we don't hear about that enough.

Parker: So, what happened is we scaled out this manual process. We had a ton of people doing things manually behind the scenes, and that led to a series of problems, and one of them is it actually made it a lot harder to automate things, because once things are being done manually at scale, it's incredibly difficult to go back and automate it.

The program's too complex, it's too hard to build software to replace what people are doing.

And the second thing was that, when we were doing things manually, there's always an error rate around that.

Garry: What do you say to founders when you meet them who are in that situation?

Parker: People are able to achieve, accomplish so much more than they think they can, and as a founder, it's your job to push the company, and to challenge them and try and understand why can't we do this much more.

Garry: We have a growth knob, basically.

Parker: Yeah, why can't we continue turning this knob? There are certain levels that I wouldn't turn it past.

One of the mistakes we made, in our efforts to get demos and get companies interested in our software and schedule demos for our sales team, those started to slow down. And one of the mistakes we made as we looked at that, and we said, well, it's twice as hard to schedule a demo now, so let's do twice as much of whatever we're doing to get the number; now we need to hire twice as many SDRs to schedule the same number of demos.

And that was a mistake. Once things started to bend like that, we should've pulled back and said, this is actually an indication that something's going wrong and we need to fix and address those underlying issues.

And look, one of the perils of growing really quickly is all of this happened in the space of six months.

Garry: Absolutely. The blink of an eye.

Parker: Yeah, it was like things started to slow down, and seven months later, I was gone.

So, there wasn't a lot of time to course-correct or adjust, but that's, I think, one of the things that we should've done. The other red flag for us was that the gross margins at Zenefits started getting really hairy.

We had all these people doing things manually, and one of the other downsides to that, in addition to the ones we mentioned, is that you have just many fewer degrees of freedom if things start to slow down.

The cost structure is so high.

Garry: And that's one of the things that often, in hyper-growth, especially past the Series B, growth investors really want gross margins to expand because of network effects or other things, so if it goes the other way, then you have a double whammy in that next round.

Parker: We didn't really get questions about gross margins in our Series A, and really even our Series B.

And then, suddenly, in Zenefits' Series C, it was the only thing that people were focused on.

Garry: So, think about this earlier, that's a pretty key thing.

Parker: Yeah, so, that was actually one of things with Rippling, the sort of central thing that we did differently is that we decided from day one there were gonna be no ops inside the company.

We weren't gonna do anything manually, and that meant that it had to be software end-to-end. We had to do the hard work of building software to connect to insurance companies, to do our own in-house payroll system, all these kinds of things.

The advantage of that, though, was that it meant that it was software end-to-end, it meant that the error rate is nonexistent. It doesn't mean that we don't make mistakes, but software errors, you make them once, and then you fix it, and they don't happen again. Human errors recur continuously. 

It means that the financial footing of the company is much stronger. 

We took it to an extreme early on with the company, where, for a long time, we didn't even have customer support. We were just doing that, I was doing that myself, a lot of the engineering team was doing support, which meant, of course, that the product got better really quickly.

Garry: With Rippling, you've been able to build it on bedrock, which is very powerful.

Parker: Yeah, look, I wouldn't have been able to do that if it weren't for, quite frankly, firms like Initialized that invested a big chunk of capital in us upfront, because Rippling was, for two years, it was me and 40 engineers and no one else and basically very little revenue, and we were just building. We had this view from day one that we needed to build.

The right answer was actually, it wasn't that we did too much at Zenefits; it was that we needed to do more, and so we needed to rebuild everything, an all-in-one payroll benefits and HR system, like Zenefits, like Gusto. We needed to build it in a way that it worked for large companies and not just really small ones, and we needed to build the entirety of Okta and 1Password and what those guys did and this whole device management system, and that we would tackle this employee record across the entire company.

We knew that was gonna be really hard, we knew it was gonna take a long time, but we thought, I think we were pretty sure, that if we could build all of that in one system, that system would be different, it would be better in a really fundamental and foundational way.

Not just incrementally better, but 10-x better. If we could make that work, then we'd win, and we'd win really big.

It was only possible because we raised a seed round that you guys were the first money in, that ended up being a $17 million seed round.It was a massively large upfront investment before we had really any product traction, and that allowed us to really spend the time to build this and build it right.

Garry: In that respect, in a way, almost a moon shot, but you knew where the moon was, and in fact, you were halfway there last time. It's almost the opposite of your first startup experience. You knew the product market fit was attainable, and you just had to get there.

Parker: So, Rippling was different than all my other companies in that, if I was to write out the spec on day zero of exactly what we were gonna build and where we are today, it hasn't moved, three years later, it's moved almost not at all. And I think that's not true for most start, certainly not true for my other companies, I think not true for most companies.

And the reason was that I think I really saw what worked at Zenefits, and I think a lot of people misunderstood why Zenefits worked when it was working, because people thought Zenefits was online health insurance,

Really, you need one system that handles everything.

It's a more foundational piece of business software that manages one employee record across every business system in your company.

Garry: I love that. That's one of the biggest problems that a lot of founders face. They actually put the engineering ahead of the business need, ahead of the actual problem people have, and in the end, you're a solution in search of a problem, and that's the opposite of how you've been able to build Rippling.

Parker: Yeah. Well, as you know, that's the YC way, starting with a problem that you have and looking for the right way to solve it rather than technology and searching for what it can solve.

Garry: Make something people want, do un-scalable things, and then the big lesson here is part two, actually scale it, which you've been able to do.

Thanks for reading part 1 of my interview with Parker Conrad. Part 2 is here.

You should click subscribe below to get my future interviews and posts by email. You can also subscribe and turn on notifications for my posts on YouTube here. Finally, you should follow Parker Conrad on Twitter at @parkerconrad.