I have walked into every PM job with a backlog of 200 to 2,000 items waiting for me. I have tried to groom my way through them, and I have failed every time. A 400-item Jira backlog is not a plan. It is a graveyard pretending to be an inventory. The anti-backlog is the demolition and the replacement.
What is broken
The backlog is a memory nobody reads. Items written years ago by people no longer at the company, requested by customers who have churned, on top of products that no longer exist in their original form. It rewards capture over judgment: PMs get praised for writing it down rather than for deciding what matters. And it distorts conversations, because "it's in the backlog, just not prioritized" implies "yes, eventually" when the honest answer is "no, never." The backlog is the lie that lets you avoid the no.
The three replacements
The live queue is small, capped at two weeks of work for the team, and signal-fed. Items enter from four sources only: a top customer signal cluster, an eval regression, an incident action item, or a cost spike. Each engineer's coming two weeks has explicit slots, and the queue empties. Items get done within their window or get kicked back to the signal source. If signal still says they matter, they re-enter.
Ideas go to three other places instead of the backlog. The hypothesis library holds half-formed ideas written as hypotheses, not features ("I think users would benefit from X because Y"). The signal stream catches what customers actually say and cluster-bubbles a pattern when five customers repeat it. And the kill list records decisions you have already made not to do something, so the same idea does not get relitigated every quarter. Three docs replace the backlog, and none grows unbounded.
How to migrate
Do not try to triage a 2,000-item backlog. You will fail. Instead: Week 1, archive everything older than six months in a single bulk move. Week 2, sort the rest by signal evidence and keep items with active customer signal in the last 30 days. Week 3, stop adding to the old backlog entirely. Week 4, delete it or make it read-only. The discomfort is highest in week four. Hold the line. It passes.
This week, open your backlog, sort by date created, filter to anything older than 12 months, and bulk-archive it. Then tell your team and watch what happens. Spoiler: nothing. For the next two weeks, add nothing to the backlog and see how much of what you would have added actually mattered. Usually about 10 percent.