On January 1st 2013 I posted Pitchpigeon on Hacker News (actually back then it was called Peterpings). For most people this was the first time they had seen the service. It hit the front page with 35 points and a nice thread of comments.
This wasn't just a landing page MVP with a signup leading to a black void of nothingness. It was a working product that did what it promised: delivered your iOS app pitch to over 100 different iOS app reviewers. Funny to think that a month later we now support over 200 different publishers and also do Web and Android apps too.
But I digress. The launch was a success for me and my team in that we validated our value proposition. How did we validate? We charged money for our service from day one. And people paid.
Hacker News sent around 8000 unique visitors on the days surrounding the launch, which converted into around 250 signups. We were pleased with this considering the landing page was un-optimized at this point. We attribute the good conversion rate to sites like Hacker News being right in our target niche - hackers and startups who want to get more media coverage for their apps. We got a stream of signups for the rest of the month via other tech blogs that had written about us plus residual Hacker News traffic.
Since ours is a paid service that fulfills a specific need, we knew we weren't going to get thousands of signups - and that's ok. As long as you charge real money, you only need a few hundred people to fall in love with your service before you're ramen profitable.
Unfortunately, we forgot to add tracking to our payment events otherwise you'd see a big spike in payments corresponding with the above signups graph. See my 10 lessons learned below...
I've redacted the actual numbers, but I'm happy to say that Pitchpigeon is off to a good start with 4-figure ramen-profitable revenue in its first month. In retrospect the revenue would have been higher if Pitchpigeon had supported Web and Android apps at launch, as we had many requests for this that we could only fulfill later in the month after developing those features.
That said, spending time building those features before launch wouldn't have been in the spirit of Lean - we had already done more than enough by building a working service for one type of application. I wouldn't change anything about our launch timeline - learning that our value proposition was valid as soon as possible is more valuable than earning a few thousand extra dollars at launch.
10 Lessons Learned
1. If at first you don't succeed...
Here's a dirty secret. Our launch post on Hacker News was actually our second attempt. Our first attempt went nowhere. Personally I think it's fine to resubmit as long as you aren't retrying in quick succession and you change your tactics. We improved the product a little more and waited 3 weeks until we tried again.
2. Encourage Participation
Here's our unsuccessful post title:
"Auto-Notify Tech Blogs About Your New App"
Here's our successful post title:
"My company's first SaaS app. Thoughts?"
The call for genuine feedback got a better result than the post that was merely trying to sell what the app does. People are not browsing Hacker News to be sold to - they are there to participate in a dialogue. I should have known better.
3. Time Your Post
We posted to Hacker News around morning time San Francisco. We think this is a good general timing. The core of Hacker News (YC and their peers) are having their breakfast and checking links. The rest of the US is in the middle of their work day. Those in Europe are at the end of their work day but probably still on the computer, and night owls in Asia are looking for things to check out.
4. Be as Static as Possible
I've seen a number of sites crash from being "front paged" over the last couple of years. Why chance it - make your landing page as static as possible and serve all assets through a CDN. We had no traffic-related issues, despite our stack being Heroku & Rails which right now is under criticism for queueing and latency issues.
5. Double Check Your Tracking!
Although we had mostly set up our goal funnels and events in Google Analytics, we made some last-minute changes and as a result our payments weren't tracked properly at launch. It's not a huge deal but it is a minor frustration not having all the numbers in Google Analytics. Worst case scenario we can always harvest the data direct from our database.
6. Charge Money
Charging real money for your service is the purest form of validating your value proposition. If people buy what you're selling, your service has value. We don't even have a trial for Pitchpigeon (by nature we can't, really). It's sign up for free, then if you like what you see, you pay. It has worked well so far.
7. Join the Conversation
I was personally responding to a bunch of emails and comments on Hacker News well into the wee hours of the morning (in my timezone). This is all just a hygiene factor though, you must be involved in conversations with your customers if you ever hope to learn from them.
8. Make it Obvious Where to Give Feedback
I mean even more obvious than one of those "feedback" tabs on the side of the site (which personally I never use). Our approach? We send an autoresponder a short while after a user signs up simply saying if you have any questions, just reply to this email - we were pretty amazed at the amount of responses we got from this. It was one of the primary ways in which our users told us that they wanted Android and Web app support. That helped shape our product roadmap to today, where we support iOS, Android and Web Apps and push pitches to over 200 publishers.
9. Run an Experiment
If I could go back and do one thing differently, it would be to set up some content experiments before posting on Hacker News. It's really not very often that you get a huge (and therefore statistically significant) and targeted influx of traffic for free. I should have set up some A/B tests on those 8000 unique visitors - it would have helped us make some landing page optimizations very quickly, that could then be applied to the rest of the traffic from the month.
10. Your Hacker News Launch is Fleeting
Remember that after the initial spike dies down and your post drops off the front page, that's pretty much it. Residual traffic will trickle in, but it's time to get back to work on your product! (and maybe share your lessons learned...)
Hope those 10 lessons help you with your future Hacker News Launch!
Good Luck :)