Running a weekly backlog review that actually shrinks the backlog
Backlog is not a number to be eliminated. It is a queue to be managed, and the weekly meeting is where that either happens or quietly does not.
Two things go wrong with backlog. Either nobody looks at it, and it becomes an archive of work that will never be done. Or somebody looks at the total every week, declares it too high, and nothing changes because a total is not actionable.
A backlog is healthy when work enters, ages predictably, and leaves. What matters is the ageing, not the count.
Read age, not size
A hundred work orders with nothing older than three weeks is a functioning queue. Forty work orders of which twelve are eight months old is a failing one, even though the number is smaller. Sort by age at every review and the diagnosis makes itself.
Thirty minutes, four segments
- Overdue on critical assets: first, every week, no exceptions. If this list is not empty the rest of the meeting is secondary.
- Aged work: anything past a threshold you set once. Each item gets a decision: schedule it, unblock it, or cancel it.
- Blocked work: what it is waiting on. Parts, access, approval, or a decision nobody has made.
- New priorities: what arrived this week that changes the plan.
Half an hour is enough for all four if the agenda is fixed and the data is already sorted. Meetings that run long are usually meetings that are re-diagnosing individual jobs.
Someone must be able to cancel
Most backlogs contain work that everyone privately knows will never be done: the asset was replaced, the fault was fixed incidentally, the request was duplicated. If nobody has the authority to cancel, that work accumulates and the whole list loses credibility.
Name the person. Record the reason. Cancelling is a legitimate outcome, and a queue that is honestly cancelled is trusted by the people working from it.
Blocked work is the most useful segment
The blocked list is your operation telling you where the constraint is. Consistently blocked on parts means the stock policy is wrong for the critical tier. Consistently blocked on approval means the threshold is set too low. Consistently blocked on access means the planning conversation with operations is not happening.
Three weeks of blocking reasons will tell you more about the maintenance function than any KPI on the board pack.
One chart, on the wall
Work orders by age band (under two weeks, two to eight weeks, over eight weeks) plotted weekly. It shows whether the oldest band is growing, which is the only backlog question that matters. It fits on one slide and it needs no explanation.
If the oldest band is growing, the backlog is not a queue any more. It is a list of things you have decided not to do without deciding it.
Related reading
Run this workflow on a system built for it
Every step in this guide maps to something SaharaDesk does out of the box. Book a demo and walk it against your own assets.