The roller-coaster of being Product Hunt-ed

On the morning of September 22nd I woke up to a pleasant surprise: one of my side projects (Doorbell) had been submitted to Product Hunt.

Doorbell is a free in-app feedback gathering tool that I’ve been working on for just over a year. And since I haven’t been making many efforts to market it, it’s been pretty quiet until now.

Having been a fan of Product Hunt since Ryan Hoover launched it, I knew that this had the potential of being huge for Doorbell.

So here I’ll describe the roller coaster of an experience I’ve had with Product Hunt, which started about 10 days ago…

Day 1

Tom Maseiro was kind enough to let me know that Doorbell had been submitted, and gave me the opportunity to jump in and start answering questions on PH.

I was amazingly surprised and humbled to see that Doorbell was staying near the top of the rankings, without me trying to game the system. OK, I did post the obligatory shameless self-promotion tweet mentioning that Doorbell was posted on ProductHunt, but most of the votes came from people who had not heard of it before.

Between people posting comments on ProductHunt, receiving a lot of tweets, and submitting feedback on Doorbell itself, I got enough suggestions to overflow my Trello boards!

The numbers of day 1

Google Analytics reports:

  • 2,225 sessions, with a previous maximum of ~50
  • 3,841 page views, with a previous maximum of ~350

And from Doorbell’s stats:

  • 59 signups, previous maximum was 5

Takeaways from Day 1

  • The conversion rate was about 2.65%, which means that some landing page optimization would probably be good.

Day 2

Highs

The following day started much slower, as expected. Doorbell was no longer near the top of PH’s homepage, since the new day’s hunts had started being posted.

Although it was a slow start to the day, I was hopeful that Doorbell might make it into the top hunts of the day newsletter. And sure enough, it made it!

Which in itself lead to an even better day (for signups at least)!

Lows

However, things took a bad turn on day 2.

In the evening, one of the services in my stack (Pound) decided to stop working. Luckily I had heard from a friend of mine that I could use Nginx instead for the same thing. After a stressful hour or so, I managed to get it working, but not without people noticing that things are a bit broken…

But, that was not the end of it.

As more traffic started coming, a security problem surfaced…

Luckily it seemed to be an edge case which only happened twice. Both users kindly reported the problem to me immediately and didn’t take advantage of it. I immediately started working on a fix, and in the meantime, not very gracefully, had to take Doorbell offline to prevent other people running into the same problem.

Fortunately, I managed to keep the feedback collection part of the application working while Doorbell was offline, so it was only a partial outage (no signups or logging in).

After about 5 hours of debugging, and nearly giving up several times (if it weren’t for the encouragement of my fiancée), I finally managed to identify the problem and fix it. The root cause of the problem (and the fix itself) probably deserves a separate blog post!

Numbers of Day 2

Despite the slow start of the day, and Doorbell being offline for 1/4 of the day, Doorbell managed to have 61 signups!

Google Analytics reports:

  • 1,899 sessions
  • 3,250 page views

which is a 3.2% conversion rate. Better than day 1, but still needs some work!

Days 3–7

As expected, the signups started dying down over the next couple of days, as did overall visits:

24th: 51 signups
25th: 24 signups
26th: 13 signups
27th: 6 signups (weekend)
28th: 4 signups (weekend)

24th: 813 sessions
25th: 345 sessions
26th: 219 sessions
27th: 117 sessions (weekend)
28th: 86 sessions (weekend)

1-Week summary

  • 5,694 sessions
  • 10,866 page views
  • 215 signups
  • 167 Doorbell applications added, 56 successfully installed (49 javascript, 7 iOS, 4 android)
  • 338 feedback threads created (173 positive, 115 neutral, 50 negative)
  • 246 replies submitted

Conclusions

Product Hunt is an amazing way to get lots of potential customers to check out your product. And getting a huge influx of visitors/signups allows you to really discover things about your product you never thought you would.

For example, a $10 DigitalOcean server coped just fine with this sudden increase in traffic. Adding Varnish to my stack and tuning the server definitely helped with that.

Lots of positive feedback and ideas were suggested by the new users. There were 39 feedback threads submitted to Doorbell, aside from the chatter on Twitter!

And probably the most important part of being Product Hunted are the new opportunities that opened up:

  1. Being listed on Product Hunt re-surfaced a conversation with Liam from Trak.io about a Doorbell-Trak.io integration. The conversation had died out almost a year ago, but now the integration is already under way! (Update: read more about the integration)
  2. Getting to write about the whole experience and publish my first post on Medium…

So thanks again to Tom for submitting Doorbell, thanks to the amazing team at Product Hunt who enable people like me to promote their products, and thanks to all the Doorbell users, new and old, for your kind words, feedback, and patience during this roller-coaster ride!