Startup diary: Just how valuable is your firm's software?
Just how valuable is the software that you've built in your startup? The answer to this question is often not very valuable at all. This may surprise you. Surely the secret sauce of your company is soaked into the very fabric of the code that your developers have so carefully woven? It is easy to be led astray by the notion that software is something of value by itself, and this mistaken belief can easily lead you into making bad decisions.
First, let me explain why startup software is not very valuable at all, and then I'll tell you how to leverage this understanding to give you and your company an edge. Let me start by excluding software that is derived from state-of-the-art research - mostly. The first implementation of Google's search engine used the now famous 'PageRank' algorithm to give much better results than other search engines at the time.
You'd think the algorithm would have been a closely guarded secret. It was not - anybody could read up on it as it was published as an academic paper.
Your own code is perhaps the least valuable code that you use - after all, it only has one user - especially in the early days. It takes a long time to discover all the bugs, and add in all the edge cases and little features that really make software valuable. It will take years. It's just not that hard to replicate two or three years' work, especially from a small startup team.
Software without people suffers from something called 'bitrot'. Technology does not stand still. Startup software is built on layers and layers of underlying technology. If you don't keep pace and upgrade, things start to break.
Software without a team to maintain it, and move it forward, is not much use to anyone. The value of a startup's technology is embedded in the team and culture that produced it. A good tech team is sustainable and carries the business forward. That's why investors care so much about the team. The whole is worth far more than the parts.
In a typical startup, most of the software is undocumented. It would take nearly as much time again to prepare it properly for use by a different team. This is as it should be; you have to move fast.
The system will only have value if the business works, and that means working fast. The trick with a startup is to document just enough to keep the team efficient, and leave the rest for later when the business has the resources to pay down the 'technical debt'.
And technical debt is the last and most significant reason your startup code has no value - all the shortcuts you had to take to go live. Somebody will have to clean up the mess, and you have to discount the value of the code quite significantly because of this.
Now that you're feeling software really isn't worth the electrons it runs on, let me suggest an approach that will make it much more valuable in the long run.
Mostly, you should give your software away. That means your company should be a good citizen in the open-source software community. The more generic parts of your code should be open-sourced.
To open-source means to release the code under a special kind of copyright licence that makes it easier for others to use your code, even for commercial purposes. Don't think for a minute that your competitors will take advantage of this; they're far too wound up in their own affairs. Using other people's code is harder than it looks. There are parts of the world that will copy everything wholesale. Guess what? You're not going to be competing in those markets anyway. At Voxgig, we open-source a great deal of our infrastructure and generic code at github.com/voxgig. Take a look to see how we've chosen appropriate material to open up. Our core business logic remains private but we increase the value of the rest by making it usable by others.
You should support the open-source frameworks that your developers use to build your system. This can be as simple as allowing your developers to submit bug fixes on company time, all the way up to sponsoring either directly or via events. Needless to say, such community focused activities will pay dividends down the road as you try to recruit a full developer team. The Googles and Facebooks of this world are already heavily invested in open-source, and use it to prove they are good places for developers to work. Even Microsoft has embraced the open-source approach.
Startups are full of paradoxes, and the lessons to be learnt from them are often only apparent when you suffer the pain of mistakes directly. Overestimating the value of your software is easy to avoid, however, and paradoxically, by opening up the crown keys, you can sparkle a little more brightly.
Metrics: This week we have 76 open issues, and 175 closed issues.