Over at Reddit, MeirDavid has a truly terrifying post on why startups fail. He presents himself as a troubleshooter and his post is based on common problems startups in extremis experience. Like everything on the Internet these days, it’s hard to know how true his story is but it does have the sound of truth. If you’ve been around for a while, you’ve doubtless seen the type of basic mistakes that he describes.
MeirDavid begins by saying he’s audited the code base of 47 failed startups and has found the same patterns in all of them. He are some of the things he found:
- 89% didn’t bother indexing their databases. I don’t know how this is possible. The whole point of databases is indexes that make it easy to access data. Otherwise, why not use a flat file?
- 76% were over provisioned. They typically were using only 13% of the resources they were paying for. One of the points of renting virtual servers is that you can spin them up as you need them. Why in the world would you have eight times the number you need?
- 68% had auth vulnerabilities. That, at least, is understandable. It’s notoriously difficult that get security right but there are, after all, out of the box solutions for that sort of thing.
- 91% had no automated tests. This is all too common and a really serious mistake. My enlightenment came when I read that Walter Bright, known for his C and C++ compilers and for the development of the D language and its compilers, has a test for every single error his compilers have ever made and that he runs those tests with every change to the compilers. That’s the sort of thing that helps you not revisit previous errors.
MeirDavid expands on these themes and the lessons to be learned from them so you should definitely take a look at his post.There’s really no reason to be making these mistakes but they are depressingly common.