From cofounder of Scribd to startup kingmaker, Jared Friedman of Y Combinator has seen the accelerator invest in innovators on the cutting edge. In this interview, Friedman spoke with Anand Thaker, CTO of Pulse Q&A, to discuss how the CTO role evolves as the startup grows and what the three hottest areas in the enterprise phase are.
Thaker: Jared, thanks for making the time to chat with us and for all your support with Pulse Q&A.
Jared Friedman: Thanks so much for having me. It has been great working with you guys.
Thaker: Going back to your Scribd days, what does it mean to be a CTO at a younger startup?
Friedman: At a young startup being CTO means you’re the guy who writes code all day, if there’s nobody else to do it. I assume that was probably your first job at Pulse.
Thaker: I can absolutely relate, because being a CTO at a younger startup is basically writing code all day with very little time for management. It’s not that different from being an engineer with some managerial duties.
Friedman: The thing about being CTO of a startup is, if the startup is growing, it means your job description changes basically every year. The first year or two your job is probably just writing code all day long. In the second year your job is to do that, and also do some sales meetings, maybe also do some investor pitches. Then hopefully you raise money and then your job becomes finding other engineers to work on this thing with you. Your job likely switches to being a full time engineer recruiter.
Then as you hire more people, it just keeps changing. At a certain point your job becomes managing a small team of engineers, then maybe you get large enough that you actually promote some people into being managers themselves, and then your job is to manage the managers.
As the company gets larger your job keeps increasing in abstraction. We keep having to be less involved in the details and more involved in helping other people do the details. That’s often hard for a lot of CTOs. I know it was really hard for me.
Every hour that you’re spending figuring out how to scale your infrastructure and keep your website from falling over is an hour that you’re not spending improving your product and getting more users.
Thaker: Scribd grew a lot in its early stages, thanks a lot to you. How do you plan for a transition at an early startup as you have to deal with the challenges of scale?
Friedman: Scribd had a lot of scaling challenges because we worked in consumer facing site where we had, unfortunately, a low ratio of revenue-to-users. So in order to get any significant amount of revenue we needed to have bagillions of users. We had eighty-five million users a month, and we weren’t making very much revenue at all. We also grew Scribd at a time when there was much less information and infrastructure… not infrastructures. There’s much less information and tools available to scale websites then there is now. When we started Scribd, Amazon hadn’t even invented EC2 yet, to give you a sense of low long it’s been.
When we were running Scribd there was a period of several years when we were the largest rail site on the internet. This is before Twitter eclipsed us as the much larger rail site on the internet. For a while we were really on the forefront of how to scale a rail’s application. It sounds fancy but what it means is that, instead of working on our product, we were spending a lot of time figuring out how to make rails scale.
Here’s the thing about scalability. Engineers love to talk about scalability because scalability is a really fun problem to think about.But as a startup your job is not to work on fun and interesting computer science problems. Your job is to actually build a business that makes money. Every hour that you’re spending figuring out how to scale your infrastructure and keep your website from falling over is an hour that you’re not spending improving your product and getting more users.
[A] large part of my job here is not working with companies, it’s actually building Y Combinator
Thaker: That’s certainly great advice and I think we’ve all heard that we shouldn’t prematurely optimize. Changing gears a little bit, after Scribd you joined Y Combinator. How was the transition from being a CTO to being a VC?
Friedman: I really like my particular job at Y Combinator because I get to work very closely with early stage startups and be tactical and hands on, rather than way out in the abstract and very far away from the place that the work is actually being done. And I also like my job at Y Combinator, because a large part of my job here is not working with companies, it’s actually building Y Combinator, which is itself a startup and has, among other things, a pretty large software team that builds a lot of really cool software for our founders and for startups. So, I spend about half my time working on all the software that Y Combinator produces.
Thaker: We’ve certainly been beneficiaries of this. We love Bookface and we’ve benefited a lot from it. There’s always interesting content on there and we met very interesting people through it. Y Combinator gets a lot of applications now and it’s increasing every year. What are three areas you think are hot right now in the enterprise phase?
Friedman: There are three areas that are hot in the enterprise phase right now. AI for enterprise is a really hot area, there’s a lot of companies that see a big opportunity in helping enterprises adopt more AI and machine learning technologies. There’s companies that are doing pure ML stuff where they will help enterprises deploy models. And then there’s also the secondary effects of that where a lot of companies that are applying ML to a specific business problem and selling a solution that’s ML-powdered. There’s lots of ML powered sales tools, ML-powered recruiting tools and ML-powered finance tools. That’s the biggest broad category of enterprise.
The next two categories are security and Dev Tools. Y Combinator has always had a really strong set of Dev Tools companies applying a lot of the really core infrastructure now for development like Heroku, Stripe, Mixpanel, PagerDuty, Doker and CoreOS. Those are all YC companies. I think we’ve gotten so many great Dev Tools companies because Paul Graham, Y Combinator’s founder, attracted a lot of people who were extremely technical and also cared a lot about programming.
We also run Hacker News, which is the largest community of software engineers, and so we’ve long had a lot of Dev Tools and security stuff applying to YC, and we continued to get a lot of that.
CIOs also often want to find out what the latest stuff is because there might be a perfect solution to a problem that they have. But there’s a little bit of market dysfunction where it’s hard for them to find each other.
One thing that has changed is that a lot of them now are going after enterprise customers much earlier than they did in the past. In the past they’d ordinarily start with indie hacker types, like developers just hacking on stuff and it would take a long time to get to the enterprise. Whereas now, people are figuring how to get to the enterprise a lot sooner.
Thaker: In the CIO world there’s a trend for large company CIOs to engage with small companies that are using the latest, leading edge technologies. How do you think they should engage with these companies?
Friedman: That’s actually something that I would love to see Pulse help with. It seems to me from my perspective, that there’s a little bit of a market dysfunction happening. Where early stage startups are building tools for CIOs, obviously they want to get in touch with CIOs to sell their product. And CIOs also often want to find out what the latest stuff is because there might be a perfect solution to a problem that they have. But there’s a little bit of market dysfunction where it’s hard for them to find each other. So I hope that Pulse can help with that, I think that would be really powerful.
Thaker: This has been great Jared. I really appreciate you for being so generous with your time. I’m sure our community will enjoy this conversation and your insights.
Friedman: Thank you.