Build it or buy it?

So you’ve decided to build it yourself instead of buying it. It’s not your core offering, but it’ll be custom and fancy and you get infinite flexibility, where off-the-shelf offerings are much more generic. Plus it sounds fun!

What’s that going to cost you?

Well, you’ll need to actually build it, so that’ll take that much time or money, plus probably some overrun if your forecasts aren’t practiced or if you hire help that runs into problems. At some point, though, that’s done! The cost is sunk and you don’t have to think about this anymore.

Then you’ll need to integrate it with all your other stuff. Then you’ll need to maintain it, and also those integrations, and when the shape of something your not-as-newly-built thing connects to changes, you’ll need to change too. Plus there’s security vulnerabilities and maybe infrastructure, and possibly even your regulatory environment.

How much time of the year will that take? How many people, or hours of your own, will need to be dedicated to it? What could they do instead?

There’s a line of answers here that lead clearly to “yes, build it”, and if you’re getting there from here, by all means, fire up git init or your equivalent new box to build in. But maybe you should just buy it.