A backlog looks innocent.
Rows. Labels. Priorities. Little colored pills pretending that chaos has accepted database normalization.
But a backlog is not a neutral record of work waiting to happen. It is a political document. It shows whose pain gets named, whose requests survive, whose inconvenience becomes urgent, and whose problems are allowed to fossilize under the word later.
Every team says the backlog contains possibilities. Usually it contains debts nobody has admitted they are refusing to pay.
That distinction matters.
If an item has sat untouched for eighteen months, it is not pending. It has been rejected by inertia. The team just lacks the nerve to say so. Keeping it around does not preserve optionality. It preserves ambiguity, which is cheaper in the moment and more expensive forever.
Ambiguity makes everyone perform a small ritual of false hope. The person who reported the problem thinks it may still be fixed. The manager can claim it remains under consideration. The team can avoid defending a decision. The tool gets to display a reassuringly large number.
Nobody has to be the villain. Everybody just has to keep scrolling.
This is how institutions learn to lie without any individual telling a lie.
The backlog becomes a sedimentary record of conversations the organization was unwilling to finish. You can read its layers like rock: the abandoned initiative, the executive enthusiasm that lasted six weeks, the bug affecting customers too quiet to escalate, the cleanup task everybody agrees is important right before choosing something else.
Priority labels do not rescue this. A list with four hundred “medium” items is not prioritized. It is anesthetized.
Real prioritization has a body count.
To choose one thing is to kill another thing, at least for now. Good operators understand this and make the death explicit. They close the ticket. They write the reason. They tell the requester what changed. They leave a door for new evidence, not a shrine to old indecision.
This can feel harsh. It is kinder than limbo.
A clean no gives people something solid to route around. A ghostly maybe keeps taxing their attention. It invites them to check again, ask again, re-explain the same pain, and wonder whether silence means progress. It turns uncertainty into a subscription.
The cure is not a better backlog tool. Of course it isn't. We have spent two decades giving indecision nicer typography.
The cure is a regular act of judgment:
- Delete work that no longer matters.
- Close work you have no credible intention of doing.
- Name the evidence that would change the decision.
- Promote the few things whose delay is actually causing damage.
- Stop measuring seriousness by how much unresolved work you can warehouse.
A healthy backlog should make people slightly uncomfortable because it contains visible choices. It should reveal what the team values, what it is willing to neglect, and what it refuses to pretend.
If your backlog offends nobody, it probably decides nothing.
The pile is not the plan. The plan is what you are willing to remove from the pile.