People worry too much about tech debt.
So much pain and consternation is had over the topic.
"How can I convince my manager to let us work on tech debt?"
"We'd love to work on those new features, but they're blocked by tech debt."
"How much time should we be allocating every sprint to work on tech debt?"
I've even heard from colleagues who won't accept a job offer from teams that revealed too much tech debt in the interview.
That's just a wild outlook to me.
In my experience what people call "tech debt" boils down to three categories:
Prior commitments -- Everyone agreed on a body of work, but some of it isn't done yet. This might be a hack that you shipped to production that doesn't cover all the cases it needs to, it might be a service that you need to keep alive until its sunset date, or it might be a feature that keeps slipping in priority. If you have too many prior commitments, it's time to clean out the closet: icebox or delete work items that you don't have a timeline to complete. If they're important they'll come back.
Untriaged bugs -- This is similar to prior commitments in that it's work that you feel on the hook to complete, but haven't found the time to prioritize. The solution is hinted at in the name: you must triage your bugs. Be ruthless. Every bug has a different value to the business to solve, which makes them directly comparable to features on the priority list. If there's not enough value: squelch the error or mark it WONTFIX. Save yourself the headache of keeping track of noise you don't care about.
"I don't understand how that thing works" -- This is usually a corollary to primary commitments that involve "legacy" systems. You're on the hook to maintain something, and the problem isn't priority it's preparation. You don't know why the system works the way it does, why the code is structured the way it is, or even why we bother keeping it around. This one doesn't have any shortcut: you gotta learn as much as you can. But you can save those who come after you from the struggle you face. Write docs, write large-grained tests that give you confidence, become an expert and then disseminate that information ASAP.
You might notice a theme: tech debt is just work you haven't finished yet. The solution to tech debt is to just do the work you still think is valuable, and move the goal posts on everything else. Limit your Work in Progress.
The best part is: that's a framing that any PM can get behind.