I fully believe in testing, but as with all things in life, you have to consider the ROI. I work on one project where we write lots of automated tests, which of

I fully believe in testing, but as with all things in life, you have to consider the ROI. I work on one project where we write lots of automated tests, which often take 2-3x as long as the initial development time, and easily outnumbers tbe SLOC 10:1. Even whith that our code coverage is maybe 50%. I would say our product is pretty damn stable and very few bugs are reported. I also work on another project run by clueless "analysts" with NO tests at all. Guess what? Shit breaks all tbe time. But that could be because we also don't have requirements, any kind of design, peer review, repository, much less anything resembling CI. We fix everything in production and everything is pretty much a failure but hey it keeps me employed for a long time.