Production planning spent two decades fighting staleness and won. The batch-weekly plan that was obsolete by the morning meeting gave way to continuous replanning, and plans now update as reality does. The victory is real. So is the bill, and the bill has no line item because nobody measures it.
The bill is churn. A system that can replan whenever anything changes will replan whenever anything changes, and every replan moves the ground the factory stands on. Setups get resequenced after materials were staged. Orders are promoted, demoted and promoted again within a shift. Planners call this nervousness, and the word is well chosen. The system reacts to every stimulus with total sincerity and no judgement about which stimuli deserve a reaction.
The cost the consensus ignores
Churn costs arrive through three doors. The first is physical. The software counts setups and changeovers, but not the cost of re-deciding, re-staging and re-communicating a change, so it happily trades a hundred small disruptions for a marginal theoretical gain. The second is material (e.g. expedites reversed, and stock repositioned toward a plan that no longer exists by the time it arrives). The third is the expensive one, trust. A shop floor that has watched the schedule change three times before lunch learns to wait, or worse, runs the informal plan kept in the spreadsheet where thinking still happens. At that point the official plan is a suggestion.
Replan when it pays, not when it is possible.
Instability, measured
The strange part of the consensus is that nervousness is perfectly measurable, with metrics that are not exotic.
- Schedule churn. Count the fraction of released decisions changed in the next release, weighted by how close to execution they were.
- Action reversal rate. Count how often an instruction was countermanded before it completed. That reversal is the purest waste in the system.
- Plan half-life. Measure how long a released schedule survives before half its near-term decisions have moved.
- Nervousness per cause. Trace which signal classes trigger the replans that end up reversed. That trace turns damping from a philosophy into an engineering task.
No planning vendor publishes any of these for its own product. Churn sits in the seller's blind spot because responsiveness is the feature being sold.
Damping, with the trade-offs stated
Every damping tool carries an honest cost. A frozen zone buys stability by refusing late improvements close to execution. Threshold-gated replans fire only when the projected gain clears the cost of the disruption, which means someone has to model that cost at all. Batching minor changes trades minutes of delay for fewer reversals. None of these is free, which is exactly why the instability metrics must exist first. You cannot tune a damper against a cost you refuse to measure.
Horizon plans with the damping built in. Replans fire when the improvement clears a stated threshold, stability metrics ship with the schedule, and the trade-off between responsiveness and churn becomes a dial you set on purpose. Publishing the nervousness numbers is the discipline the category avoids, which is what makes it worth buying. Start here.