Small departures from procedure that favor a new strategy are allowed one at a time until they become the way work is done, and formal approval never happens.
An example
At an orthopedic implant maker, the operations team was pushing to move sterilization to a contract facility to cut lead times. The validation procedure required three full qualification runs before any product could ship from a new site. The first exception allowed shipment after two runs for a backordered knee system, approved by the plant director as a one-off. The next quarter a deviation let the contract site use its own load configurations. Then a run that failed on paperwork was accepted with a note. Each exception was small enough that the change control board never saw it. Reviewing the file ahead of a regulatory audit, a quality engineer found that the contract site's practice no longer matched the written procedure in four places, with no approved change record for any of them. The engineer wrote up all four as a single change request and routed it to the board.
Why it works
Small deviations look easy to let pass: no committee wants to fight over an edge case. As deviations pile up, the formal procedure looks worse, since everyone can see it no longer describes how work is done. Reversing looks costlier over time, because people have built workflows and trained successors around the new norm, and rolling back exposes the leaders who tolerated it. It feeds on governance with too little bandwidth to flag small deviations, a pace that beds in the new norm before review catches up, and a culture that treats minor exceptions as judgment calls.
The thinking behind it
Leon Festinger Festinger's A Theory of Cognitive Dissonance (1957) holds that people reduce the discomfort of a belief at odds with their actions, often by changing the belief. Engineers who sign off each deviation adjust their sense of what the procedure requires to match what they approved, and the written procedure starts to feel out of date. Aggregating the deviations into one decision makes the gap between rule and practice too large to absorb that way.
Richard H. Thaler Thaler and Sunstein's Nudge: Improving Decisions About Health, Wealth, and Happiness (2008) calls defaults the strongest feature of choice architecture, because whatever happens when people do nothing is what most people end up with. Once a deviation is accepted, continuing it is the default and reversing it takes action. A forcing function such as an audit turns the cumulative deviation into a choice someone must make.
Warning signs
Current practice diverges from the written procedure, with no approved change record. Each exception was too small for any committee to fight, and many were approved as one-offs. The deviations all lean in the same direction, toward a particular strategy. Staff describe the procedure as out of date. New people learn the practice, and the document becomes a formality.
How to defend yourself
Add them up. Individual deviations lose on their own merits, so stop arguing them one at a time. If the new practice is good, the single decision will approve it and the procedure will catch up. If it is not, the organization finds out before an incident shows it.
- Compare current practice with the written procedure, and list every divergence with its date and approver.
- Package the list as a single change request, so governance must approve or reverse the cumulative change as one decision.
- Time it to a forcing event if one is coming, such as an audit or a quarterly review.
How a leader shuts it down
Install forcing functions, such as an outside auditor or a quarterly process review, that add accumulated deviations back into a single decision against the original rule. Track exceptions by direction as well as by size. Treat a gap between written procedure and practice as a governance failure, whatever its cause.