Killing Kojitsu

Let’s start with a bit of context. About a year ago I started work on a little web application that could surface the best links shared by the people I follow on twitter. Last February I released that application into the world as Kojitsu.

I’m fascinated with the potential to tap the social graph and create a new type of social recommendation engine. Kojitsu started out as just an experiment to explore that fascination.

After operating Kojitsu for about six months I decided to shut it down at the beginning of July. In this post I want to share with you some of the reasons why I came to that decision, and the lessons I’ve learnt from building the product.

What Kojitsu did well

Before we dive into what went wrong, let’s take a look at what Kojitsu did well.

The number one thing that would come up time and again in chats with users was how much they liked the simplicity of the interface. I’m a big believer in the power of simplicity and worked hard to make the app as intuitive as possible so this meant a lot. As any maker will know, having someone pick out a certain aspect of a product you worked hard on brings a great sense of accomplishment.

Of course any content discovery product is only as good as the recommendations it generates. I feel that Kojitsu did a good job of hitting the mark here, and users seemed to agree. Don’t get me wrong, the ranking algorithms were far from perfect. But they were consistently surfacing interesting content.

Why Kojitsu had to die

When I shut down Kojitsu on July 11th the service had 218 registered users. The majority of these signed up when Kojitsu hit the top of Product Hunt back in April.

For a lot of people that’s a small number, but it doesn’t feel small for me. I come from a B2B world where I know the name of every single one of my customers.

When you’re focussed on statistics, it’s easy to forget about the people behind the numbers.

I genuinely care about every single one of those 218 people. They took the time to try something I built. Many of them offered me feedback. Shared Kojitsu with their friends. There’s no greater satisfaction for a maker than seeing people enjoy something you’ve built.

So where’s the problem? Having users is great but you don’t have a business unless you have a way of making money. I was bootstrapping Kojitsu with the money I was earning as a freelance writer. To begin with I was okay with that (after all I had built this to “scratch my own itch”) but as Kojitsu grew, so did the cost of running the service. When the money from my writing gigs dried up I had a tough decision to make. I needed to monetize fast, or shut down the product.

There’s a perception in the startup scene that if you build something, people will flock from afar to use your product. Then when the time is right, you can turn on the money spigot and start generating ad revenue off all those eyeballs. That business model may have worked for a handful of companies, but it’s rare.

With Kojitsu I tried a whole bunch of different ways to ‘turn on the money spigot’. I spoke to users about charging for the service, explored sponsorship deals, advertising opportunities. At one point I reached out to social media agencies to see if turning Kojitsu into a B2B service might work. I thought they might want Kojitsu as a tool for finding content to share through their clients’ social media accounts. They didn’t.

Nothing worked. There was no business model. There was no business.

So that’s why I shut down Kojitsu. It had reached the point where Kojitsu needed to grow up and become a business to survive. Without a viable business model, that just wasn’t possible.

Lessons Learned

The whole experience with Kojitsu has taught me a lot. Far too much to note down in one blog post, but I want to share with you some of the biggest lessons I’ve learnt.

Charge for your product

I’m done with this perception that stuff on the internet has to be free. Free is not going to pay your rent.

Build something awesome and charge people for it. There’s no greater validation for an idea than having someone hand over their hard earned cash.

It seems crazy I even have to write about this, but a lot of people are falling into this trap (hey, I did!).

Tell anyone outside the tech industry that you’re going to give away your product for free and they’ll laugh in your face.

Free is not a business. Free is a hobby. Charge for your product.

Don’t be afraid to cannibalize features

In the early days, one of the top feature requests was the ability to get recommendations delivered as a daily email. Kojitsu started out as web-based application so people would need to log in to view their recommendations. Building habits like that is hard, so a daily email sounded like a great idea.

I completely underestimated just how much of an impact this would have on the core product. Overnight almost everyone stopped using the web interface and switched to the daily emails.

I had cannibalized the web interface that I’d worked so hard on by launching the emails. That can be a little daunting, but the emails delivered a better experience so it was the right thing to do.

I’ve seen people work on something for days just to justify a few hours they spent in the beginning. If something’s not working out, cut your loses and a move on. You might incur a little short-term pain, but you’re securing the long-term health of the product.

Humans are awesome

Out of all the experiences I had whilst building Kojitsu, the points that stand out most are the chats I had with users.

Feedback is crucial to a young product. The Kojitsu community gave me some of the best damn feedback I’ve ever received.

It means a lot that people are willing to take the time to share their two cents. I hope I have the chance to repay the favor many times over.

Why don’t you just open source it dude?

After I announced that Kojitsu was shutting down a lot of people asked if I would open source the codebase. I decided not to. Not because I’m possessive over the code, but because I don’t have the time to maintain it.

I don’t want to just throw some code over a wall, then turn my back and head off into the sunset. Open source projects are only successful if they have a healthy community around them. Open sourcing the codebase so that it can rot out in the open wouldn’t be the right thing to do. Sometimes it’s better to just let things die.

What’s next

Shutting down a product sucks. I feel like I’m letting people down, and I hate that. But everything comes to an end at some point.

As for what’s next, I’m already deep into development on a new product (one with a real business model!). More on that to come soon.

My time with Kojitsu has taught me a lot. I hope that some of the experiences I’ve shared in this post will help you in your own endeavours.

Until next time.

Keep hustlin’


Originally published at blog.mattwest.io on August 8, 2014.