I think that my case is unusual, but I did a comp sci PhD, writing a dissertation on relational databases, and it's been extremely useful. I pursued the PhD bec

I think that my case is unusual, but I did a comp sci PhD, writing a dissertation on relational databases, and it's been extremely useful. I pursued the PhD because I liked writing software and wanted to study it more. No plans about academe or industry. Just seemed like a good idea at the time. The PhD has helped in a few ways: - I've been doing database-related startups for 20+ years, and my deep understanding of database technology has proven very useful. I could have gained some of this insight by working, but I think this path has led to a broader and stronger foundation. - On a couple of projects, my actual dissertation subject has been relevant. - While a PhD was off-putting to employers at my first startup (understandable, since it was also my first actual software job, and they had no evidence that I actually wanted to write software and was capable of doing so); once I got past that, the title was occasionally a useful thing to put out there. (On one occasion, I was asked to write a paper for an industry group with "lots of symbols".) - Contacts with the academic world of databases that have come in handy over the years.