Sell shovels, not homes
When you walk into a Home Depot, you see everything a carpenter needs to build a home from scratch. But one thing they don’t sell? A pre-made home.
There’s a growing trend of developers who are starting PaaS (platform as a service) companies, and it’s alarming to me how quick developers jump at the opportunity to build services for other developers. As Slava Akhmechet points out in his excellent RethinkDB post mortem:
Developers love building developer tools, often for free. So while there is massive demand, the supply vastly outstrips it. This drives the number of alternatives up, and the prices down to zero.
But here’s where the opportunity is: developer tools can be a profitable business, you just have to sell them shovels, not the house.
Turn technology into a distribution channel, not your product
There are at least three startups right now creating PaaS platforms targeted at developers who don’t want to run their own backend . They bundle and resell really hot technologies, like GraphQL, Lambda, and microservice architectures, and combine them in ways so developers don’t have to worry about how they are interconnected and scaled. But the problem is, developers are the same as carpenters who are building homes. They don’t want someone to build it for them. Developers are builders that want to leverage the best tools to build their own house.
However, by viewing said technologies as channels instead of products, startups can sell tools and services that make serious money.
For historical precedent, here are some companies that come to mind that took this approach:
- JBoss — a collection of Java middleware that leveraged SOA principles into building blocks that developers can use for their own stack (like plumbing).  
- New Relic — leveraged instrumentation capabilities in .NET and Java to help developers monitor performance of their stack (like electricity metering).
These are just two companies that leverage the technology space to offer tools that supercharge a developer’s productivity, without building the house for them.
GraphQL as a channel
It will be really interesting to see how companies build on top of GraphQL. There are some really amazing use cases that can be unlocked when developers have a common interchange format, which they can use to simplify the work between the frontend and backend. But I believe the most interesting developer facing companies of the future will be ones that leverage GraphQL (and other new technologies) to break down the data silos between organizations, and who make it easier to work with the tech — not build their entire house for them. 
 — I believe one of the reasons we’re seeing more PaaS out there is because it’s at least 10x easier to build with AWS. But that doesn’t mean it’s the best business to start.
 — Does anyone else remember SOA and how WSDLs were an early precursor to GraphQL?
 — Definitely read @BostonVC’s excellent article on how JBoss rose to the top
 — http://www.apollodata.com/ looks like the leader so far here, but I’m sure there are at least 10 other GraphQL related tools that could turn into big businesses.