The buckets-based triage from yesterday can work, but it's still not my preferred way to replace a backlog.
And neither is this one, but here goes.
Instead of a backlog of accessibility issues, only keep three accessibility priorities at any time.
When one's done, you find the next. That's it. You're not allowed to have a fourth.
There's no list behind this technique. These three things are what you're actively working on and it's safe to say you can easily keep them in your head. If you're stopped on the street and someone asks you what are you working on, these three things will instantly come out of your mouth.
So that's easy. The hard part isn't the system, but rather how you choose. Three is a small number. It will force you to have a real conversation about what matters right now and that's uncomfortable, because it means saying no to everything else. That discomfort is the point. A backlog lets you avoid that conversation altogether by just adding another ticket.
This works really well alongside the fix-on-touch approach. Your three priorities are the things you're deliberately going after and everything else gets fixed when someone's already in that code.
The obvious downside is that things outside your three become invisible. If a critical issue pops up, you should ignore it until you've fixed at least one of your current priorities. Otherwise, you're suddenly reshuffling, which is what you wanted to avoid.
The real value here is the constraint. If you have capacity, work on five issues at a time. Six. 20. Pick your own limit. The limit is what matters because a limit forces decisions and decisions are what move accessibility forward.