Earlier quoted context omitted.
>>>How? Programmers don't have executive powers. The secret is - and this is wielded a million times a day in a million small ways - of course we do. Lots of people do in lots of jobs, whether intended or not. You just do the thing . You don't ask about doing the thing, you don't put in a project planning proposal, you don't beg your manager to devote 20% of your time to the thing, you just go and do it because you p…
When do you do it? In between the constant firefighting or in between poorly planned “we need this yesterday” tickets?
For example, is there some process improvement such as getting more data, new monitoring tool or logging solution that would reduce time spent in firefighting? Something that would immediately help you solve the next issue more easily? Do it while working the issue.
Can you rewrite the tickets so they are no longer poorly planned, making them easier and quicker to implement? Do it while working on one ticket. Depending on the organization, you can probably rewrite the tickets even if it is not your job, but that may cause waves. If that is too risky, another alternative is to create sub-tasks for the high-level issue that are better planned. If you need another high-level ticket, just ask for it.
When work is not optimal due to time pressure, there are probably costs related with that. Saving time and resources requires longer-term perspective and making some investment. But often there are things that require only smallish investment, and that is what you can very likely do.
When you save time for your company, re-invest that time into some other improvements that bring bigger results but might require a little bit more effort.
The lesson is that you have more flexibility than you realize, because your manager probably does not understand what you are doing. If you get good results, you may be eventually granted more flexibility. But that will most likely take time if you are working in an environment that is constantly fighting fires.