I think my biggest failure this time was getting emotionally attached. Yes, the project did not generate any money, but it did what it was supposed to do - (in)validate the market. Emotion was what kept me hooked on it and not letting me move on to the next thing. Even though it's sad, I'm shutting it down.

The project

Me and a few friends needed to maintain a few text-based websites and since no such service existed I jumped to conclusions and made the service happen.

What went wrong

Releasing on Hacker News drew in a huge crowd of 30'000 people, but only 2 people found it useful enough to buy. Even though the forum has mostly tech people in it (who can run up their own servers) it's still pretty conclusive evidence that such a service is not actually needed. If not that then hitting all kinds of people up on the internet is.

Instead of realising what had just happened I thought "well so many people came to see something, so there must be something here" when in reality "Markdown website in 5-minutes" just sounds really cool and that's about the depth of it.

So I lingered there for a few months fixing some crucial bugs (site is now SSL only for example) and tried contacting people and trying to make sales basically. No actual sales were made. Double conclusive evidence I would say.

But no, I couldn't give up now! I went on to ask for advice from mentor-like people and spawning some topics on Reddit to get conversations going. Well, the conversations were the ones that really struck home for me (Reddit/HN is awesome). Point was if anyone anywhere needed a website of any kind they either had a friend for it or a million other services that do the job just fine plus more.

I solved my own problem and no one else's. I'm sure there are people like me out there with the same need, but really it's very much a nice tool than a must have.

There were many ideas I could have pivoted towards, but none would fix my problem of not fixing some specific groups specific problem. I even tried creating a text-only website movement, but no one cared.

This guy on reddit sums my mistake up rather nicely. STOP building things and START TALKING TO CUSTOMERS. I started with it too late.

What went right

Didn't spend too much time on the MVP. Granted, two months is a long time just to test the waters, but it could have been a lot worse. Also, the actual product only took two weekends, it was the damn payment system that took forever (both for technical and paperwork issues).

Didn't act reactively on peoples advice. Most of it was "needs feature X" and "UX is broken" and "front page is missing", when in reality my problems were much more deep.

Making the customer pay right off the bat filtered the "want" from the "need". I found out that while a lot of people became fans and were ready to use the service, most of them never did. One came back and bought a domain.

Releasing early and rolling with the punches. The initial HN post brought an avalanche of critics. And I'm glad it did, I learned a lot about how others would have approached it, what this forum is good at doing and how people see my product from different perspectives. Luckily I was able to distinguish what was important at my current stage. Also gaining early adopters is a sure way of validating your idea, especially if they pay. I didn't get many.

Next time

Two weekends is a good time limit. First one goes to building the main structure, second one patches the worst holes - release.

Focus on communicating the vision and not actually making the product work. As in instead of a payment system I could have just had a fake form and get the persons email when the service is actually ready (not saving the credit card info of course). Not totally sure if this is MVP enough to validate a business idea, but point is I need to be sure that there are clients actually willing to pay without putting two months into it. I've just stumbled on a very similar service and I must say they did a much better job at presenting themselves AND all they have is a landing page to test the waters, nothing else. I'm getting jealous over this, but really I'm quite exited to see how far they get.

Release often and validate specific hypothesis. There's nothing quite as down to earth as letting people rip your product apart and taking it all in. But even better is when you know what assumptions you're looking to validate - separates the trolls from the advice.

Go for a market with no established leader. So climbing to the top with quality is actually doable. A smaller market is better than one with a great competition. In this case Squarespace and some other services have everything covered to the fullest extent.

Post-mortem actions

The 3 people that did buy a domain will get a personal email with instructions on how to secure their domain and web content.

Still not sure...

..if this is the right decision, but writing about this sure helps understand it better. I'll also have something to come back to when I start thinking "why the hell did I kill that project, it was so promising". I'm sure getting a lot of traffic from HN created the promising feeling, but that's the only thing that has succeeded as far as numbers go.

In any case reading IndieHackers and startup failures it's quite clear that in order to get clients you either need a very clear understanding of a problem you're solving (there's a clear lack in a field you know well) or upon releasing got a bunch of paying customers (a.k.a. early adopters who to start talking with) of which my pool of 3 was just not enough.

Okay, enough of this. It was fun, but it's not going to work. Love you, bye! killall markdownsites.