This is my pet theory, but I suspect that dev tools also get in the way of opportunities for promotion for engineers. Sure, buying an A/B testing tool will solve the problem of needing that tool, but it will add nothing to the resume of half a dozen people trying to get a promotion or an impressive project on their resume that they can then shop around at other firms. Dev tools go against resume-driven-development at larger firms, so in some sense they're doomed to fail. They make a lot of sense at earlier stages of companies where everybody's incentives are aligned around making the company win, and that might be the only opportunity the dev tool has to capture a logo and grow with them upmarket.
I fully believe this to be true as well. Aside from possibly the endless red tape and bureaucratic hoop jumping of making a purchasing decision. "Promotion driven development" is a scourge. Endless waste of time and resources for everyone, including the company, but somehow management and leadership doesn't care. Make it look like you did lots of work instead of the right work and somehow that's what matters? I know growing up I was always told work smarter not harder, but apparently harder not smarter is what management actually wants. Personally I think they're too brain-dead to know the difference and just think hurr-durr more code mean better. It's been a pretty universal experience/observation for me.
The developers and managers alike are rationally responding to the incentives of the current market. If developing a tool no one wants will get you promoted or a chance at a higher paying job elsewhere and your manager gets credit for overseeing a hard working productive team, then why would you expect any different outcome?
