Brett van Zuiden

4 Insights from a Hacker News Launch

Or "How I Learned to Stop Worrying and Love the MVP."

"You know what, screw it, let’s launch." It was Tuesday: we just had a basic demo of filepicker.io, some friends had poked at it, and we had a long "pre-launch" to-do list. But for some crazy reason, we decided to throw caution to the wind and share our fledgling product with the world. And, because why not, we decided to do it on Thursday, in two days. We thought maybe a few people will try it, give us some feedback, and if nothing else we'll push it out again in a week.

So we spent the next 48 hours cranking away, building a landing page, registration, demos, authentication, and finally at 3pm Thursday we flipped the switch. Things turned out better than we could have dreamed.

Data from the launch

  • #1 for just 2 hours, 56 minutes, holding our own above a TC article about Evernote’s fundraising at $1bn valuation
  • 10,000+ visits in 4 hours
  • At one point we had 460 concurrent users
  • 12,000+ people used the demo over the course of the day
  • 500+ developer sign-ups
  • 5000+ files picked
  • $0 in donations

Prior to this, I had never built a site with 10k visits total, so this was one of the biggest rushes I've been ever had. With the four of us glued to Google Analytics, MixPanel, and Heroku logs while pushing out bugfixes, it was a two hour roller coaster ride. So a big thank you to the Hacker News community for making our launch so much fun.

What we learned

  1. If you're thinking of doing it, launch. If you’re not quite there yet, launch. Seriously. We launched a plug-in for developers that didn’t support Internet Explorer. But it was great, because we got fantastic feedback and a set of lead users. The HN community is very understanding about MVPs.
  2. Be ready to scale quickly – thank god for Heroku. We still got bit by a few minor things (I was out at dinner when we hit our free SendGrid cap and had about 20 minutes while we couldn't send out developer keys), but we didn't go down, which was awesome. A corollary to this lesson: keep your phone on you at all times.
  3. You want to hit #1. If you hit #1 there are all sorts of snowball effects like twitterbots and aggregators that pick up the announcements when a new post hits #1. There’s no magic bullet to getting to #1, but engaging with people who comment in a meaningful way tends to help.
  4. Remember to put in your analytics. We had forgotten to hook in MixPanel events until about 5 minutes into the launch and lost out on some of the data. Never again.