I’m trying something a bit new here.

In covering startups, we tech reporters tend to focus on three very specific times in a company’s life: the big launch day, the subsequent product announcements or pivots, and, be it that the cash runs low and the audience just never comes, the death.

I think we’re all missing something. By covering the big events and ignoring all of the magic that goes on in between, we’re tossing aside what is perhaps the most important part: the lessons learned along the way.

To address this, we’re playing with the idea of a new post series: The Debrief.

The idea is simple: two or three weeks after a company launches, we’ll sit down with its founder(s) to discuss the road traveled so far. How did the launch go? Did it live up to their expectations? What would they do differently, given the chance? A few months later, we’ll check in with them again to see what’s changed. And it’s not just for launches — if a company fails, we’ll debrief them for a retrospect on what went wrong.

I won’t debrief every company that sees the light of day — hell, I won’t even debrief most. I’ll focus on companies that we find particularly unique or interesting, or with founders that I feel have tons of knowledge to share.

Take our first Debriefee, for example. After graduating from Yale in 2005, Justin Kan went on to launch Kiko (acquired by Tucows in 2006) and Justin.tv, discover companies within his company with Justin.TV-spinoffs Socialcam and the remarkably popular TwitchTV gaming portal, and become a part-time partner at Y Combinator. This guy knows a thing or two about startups.

Just two weeks ago, Justin and his co-founders launched Exec, an on-demand, flat-rate personal assistant service built in the same vein as Uber and their towncars. We covered the launch here.

I sat down to chat with Justin about how it went.


This is your third (or fourth or fifth, depending on how you’re counting) big launch. How’d it go?

Launching a new product is always one of the scariest and most exciting moments of a startup.

When we launched Kiko, which was the first AJAX web calendar, we didn’t really do much more than put it on Reddit, half to help get us some traffic, and half to help out our friends Alexis and Steve. A few days after, we were covered by a very new blog called TechCrunch, which turned on a firehose of traffic — I think we got several hundred hits that day :). Still, this brought the site down. Verdict: not prepared to scale.

When we launched Justin.tv (the show), TechCrunch dropped our story at midnight the night before we had intended, instead of 9am, which meant that the much larger instant flood of users were greeted by a live video feed of me lying in bed (and subsequently left the appropriately biting comments). The next two weeks were filled with massive technical outages as we tried to make our mobile streaming setup work, and me frantically trying to entertain the audience by scrambling around the city. Verdict: not prepared to scale or to entertain.

When we launched Socialcam and TwitchTV, I have to say we did a pretty professional job! We had beta users using both products well before launch, and we were able to use our technical expertise from Justin.tv to scale from day one. We knew the products worked because we had built them by iterating around customer feedback for months in advance. Overall, I think these went off as smoothly as I’ve heard launches go.

Exec is a new kind of product. Because customers are only hitting the site when they want a real job done, we don’t have the same technology scaling issues that we had at Justin.tv with hundreds of thousands of concurrent users. That said, we still have the potential for massive concurrency issues, because all the threads are actual Execs doing jobs. If too many requests come in at once, we get jobs going unfulfilled and customers have a bad experience.

We had great beta customer feedback in San Francisco before launch, but we were paranoid that we would launch and get a huge flood of incoming jobs at the same time that would overwhelm the system. We took longer than ideal to make sure we were staffed up, and we also implemented an invite system so that we could let people in over the course of days instead of all at once, regulating the demand side.

Things went pretty well: We’ve been able to keep up an excellent level of customer service while fulfilling all the jobs (although at peak times, I have to admit that the team has run jobs ourselves :D). If I was to do it again, I think I’d actually launch sooner (which was the advice of the other YC partners), because if something isn’t breaking you’re probably not moving fast enough.

When you say launch sooner: how? Shorter beta? Let more people into the beta, faster? Or go nuts and just skip the beta all together?

Shorter beta. I definitely think having a beta period was critical as it allowed us to improve the product drastically before launch, based on customer feedback. There’s no reason for companies to not have a live beta with customers. Without one, you have no idea if your product is going to work.

I think we could have shaved a week or two off the beta, though. Basically you should launch as soon as you’ve learned all there is to learn from your existing customer base.

Have you had any ridiculous requests come in yet?

There have been a few fun requests that have come in. On Valentine’s Day, one of our users read instructions for creating some art work on Reddit, and asked an Exec to make an exact copy to give to his girlfriend (picture below).

A few weeks ago, another beta customer used Exec to refill his empty scooter with gas and drive it from the Mission back to his office in SOMA. Both times I was a little worried that we wouldn’t be able to fulfill the jobs, but we actually did both of them well.

How has adoption of the service been? Is it in line with your expectations? Are you able to keep the Execs busy, outside of peak hours?

Adoption has been great so far. I know everyone says that, but we’ve seen consistant daily growth in job requests. The problem has actually been that we’ve been supply constrained at the moment. (We need to hire more Execs!)

What are your plans to expand outside of SF? How do you overcome the challenges of launching a real-world service in a city you don’t call home?

We want to expand to the peninsula and the rest of the Bay Area first. After that, we’ll probably go to a single new metro area. Our plan for expansion is that, when we move to a new city, we’re going to move a part of the team there so that we can really understand what is going on from the ground floor.

You founded your first company almost 6 years ago. What’s changed for entrepreneurs since, for better or worse? 

It’s easier than ever to start a company now. Today as entrepreneurs in tech we have YC, the fundamental changes in the market for seed rounds (more angels giving better terms), easy financing through convertible notes, and the ability to get something up and running on services like Heroku, Mailgun, Pusher and other SAAS platforms. I’ve been saying this is like the golden age of Web and mobile apps.

The only hard part is that, because it is so easy to start companies, there is much more competition for talent. But in the end, I think strong leaders can attract talented engineers, designers, and product people.

This is your third time going through Y Combinator, if I’m counting right. Why go through an accelerator again? Is it a matter of habit, or is it really just that advantageous? Is participating in such a program, be it YC or any other, something you’d always recommend for new entrepreneurs?

There are a couple reasons I wanted to do YC again:

  1. I’m a part time partner at YC and wanted to make sure the other YC partners would be involved in the business Day 1, since they are an amazing resource.
  2. I wanted my cofounders to go through the experience, since they are first time founders.
  3. It’s great DNA to start a company with: It forces you to have an intense focus on the product and talking to customers during a short period of time.
  4. It’s inspiring to be part of a cohort of great companies all working very hard on their products. This makes me want to work harder and do more every day.

I think YC is valuable even to experienced entrepreneurs, which is why many YC founders come back a second time. (I think I’m the first to come back a third time.)

With multiple companies and a YC gig under your belt, what’s the one piece of advice you’d give to entrepreneurs just starting to plant their roots?

The hardest part is getting started. Once you start on the path of entrepreneurship, you remove optionality and force yourself to start learning, which is the most important thing to do. Everyone who values your company when you start it does so because they think you are going to be worth something in the future, and in order to be worth something in the future you have to grow your skill set from where you are today. The sooner you start learning (becoming a better technologist, salesman, manager, whatever it takes to get your startup to the next level), the faster you’ll become effective at executing.


This is a new format for us, so do let us know what you think. (Read: let me know how much you love it and tell me I’m awesome and give me high-fives, when you see me in public.) What would you want to see? More questions? Less? Video interviews? More photos?

[Photo Credit: miss karen on flickr]