Developers and development tools

March 13th, 2010 in Micro Blog

Tools Developers like to write development tools. It makes sense too: we’re stuck in a forest of details, building, and paining over the technical bits. But the market for developer tools is fickle, and short, and more complicated than most other software.

I have to admit that I have a long list of ideas for tools: a better text editor, better design tools, better source control UIs, simpler project management, debugging, the list is as long as my imagination and frustration with the tools I use today. And the tools I’ve used over the past 20 years.

We dream of what we do

I dream of better tools, because it’s what I do with most of my time. I design software, plan projects, and cut code. It’s no surprise that many of my ideas are about software tools. And that’s not a bad thing, there are tools that are worth building.

Developers are fickle

But how many developers like paying for developer tools? How many employers? In my time managing teams and my own development process, I’m wary of buying tools. There are a few reasons, but namely its that it’s hard to find tools that actually work. The tools are expensive, and time to learn them is even more expensive.

I do pay for tools I use. I filter them carefully, as the tool has to be both useful and maintained. Not all developers are willing to buy their tools, however, and not all employers will buy more than the minimum.

Developers are the minority

Statistically speaking, software developers are a small percentage of computer users. Software developers facing the particular problem you’re trying to solve are an even smaller number. How large is your market? How much will it cost to build and maintain the tool you’re dreaming of?

The markets are small

We’re the minority, we’re fickle, we’re cheap, and we’re fragmented into subsets of platforms, tools, and methods.

Abandoned and malnourished projects abound. I’m a Textmate user, an editor with a lot of promise (and some serious performance bugs). It’s been a solid tool, but it’s also been stagnant for longer than I’ve used it. I’ve tried other editors, and they have similar problems (BBEdit being the best maintained of the set). It’s clear that the market for tools isn’t sufficient to make for solid businesses.

And this isn’t just an OS X problem. I’m also a Windows developer, and the text editor landscape there is worse. There are dozens of free editors, and a number of pay editors, many of which are updated so rarely that they’re unworkable. On ‘nix, it’s the same landscape: lots of new, underpowered editors, and a few old (and crusty) more workable editors.

The platforms are fluid

The landscape of platforms, development, tools, and what’s relevant is fluid. Versions is a beautiful OS X tool for Subversion access. Their business is based on the popularity of Subversion, which is being eroded quickly by Git (and Mercurial). Will there be enough business to sustain them?

Not only does the landscape change, you may be competing with the platform guys themselves. I’ve seen this recently with AWS and 3rd party tools: Amazon has been developing their platform quickly, making many of their 3rd party products irrelevant. How do you predict that?

Complexity kills

On top of a tough market, developer tools can be more complex than more popular things. Now the complexity is offset by your experience (you do know what you’re doing, right?). But carefully consider how much it will cost you to build your tools.

37Signals is a good example of this: they sided on a simpler form of project management. That focus was one of the reasons they’ve been profitable, combined with the wide appeal of their tool. But if you can’t keep it simple, and it has a narrow appeal, your path to profit is much more difficult.

Think twice, then think again

My advice? Question your motives when thinking of starting a business based on developer tools. Look at the market size, and what people are willing to pay. Assume the worst, then assume the worst of that. Try to imagine if the big boys are already working on it (if you think they should be), then assume they are. And think carefully about how relevant the tools will be in a year or two. Then ask yourself: would I build this just for fun?

Most developer tools fail to make money as businesses. It’s not impossible, but it’s a small, difficult, and complex market, strongly dependent on the platforms and platform vendors.

But that’s not the end of it

There are great developer tools waiting to be made. There are profitable businesses in developer tools. Just realize your motivation as a software developer: you see tools everywhere, but is there a viable business in your ideas? There may be. Just be careful not to get lost in your blind spots.