Every working day, approximately 10,000 baby boomers retire across North America. That number comes from demographic projections, and it has held steady for the better part of a decade. In heavy industry alone - manufacturing, mining, oil and gas, utilities - the figure translates to hundreds of named experts leaving their posts each week. Most of them will never be replaced by someone with equivalent experience.
Operations leaders know this. The retirement wave is not new information. What is less understood is the nature of what leaves with each expert. It is not procedures. Procedures are documented. It is not compliance knowledge. That lives in binders and training modules. What leaves is judgment: the tacit, contextual, often contradictory knowledge that an experienced operator carries in their hands, their ears, and their memory.
The gap between procedure and practice
Consider a senior millwright who has maintained the same line of centrifugal pumps for twenty-two years. The SOP for a bearing temperature alarm says: check the temperature reading, verify the sensor, reset the alarm if the reading normalizes. That is the procedure.
What the millwright actually does is different. He listens to the pump. He checks the ambient temperature. If it is below minus ten, he knows the alarm is likely a thickened-grease signal, not a real fault. He lets the line warm for forty minutes before touching the reset. He checks the cold-start grease on the number three blower specifically, because that unit has a history of seizing when operators reset cold. None of this is in the SOP. None of it was ever written down. And if you asked the millwright to explain it, he would say something like: “You just know.”
Where the documentation falls short
Standard operating procedures are necessary. Nobody disputes that. They capture compliance, they satisfy auditors, and they provide a baseline for new hires. But they are written for ideal conditions: full staffing, daylight hours, equipment operating within spec. The gap between what the SOP prescribes and what the expert actually does on the floor - under real conditions, with real constraints - is the gap that retires with the expert.
This gap is not captured by exit interviews. Exit interviews produce generalizations. It is not captured by job shadowing, because shadowing produces observation without structure. And it is not captured by the knowledge bases and wikis that companies build in the months before an expert leaves, because those tools are designed for explicit knowledge - facts, steps, lists - not the conditional, situated judgment that makes an expert irreplaceable.
The cost is real, even if it is hard to measure
SHRM benchmarks put the average replacement cost for a senior technical employee at $180,000 or more. But that figure captures recruiting, onboarding, and lost productivity during ramp-up. It does not capture the judgment-dependent incidents that occur after the expert leaves: the cold-start bearing that seizes because the new operator followed the SOP exactly as written. The safety call that goes wrong because nobody on second shift knew the informal protocol for that particular unit. The three hours of unplanned downtime that the retired expert would have avoided with a single sentence.
These incidents are rarely attributed to knowledge loss. They appear in incident reports as operator error, equipment failure, or procedural non-compliance. The connection to the departed expert is invisible in the data, which is precisely why nobody is measuring it.
What structured capture looks like
The alternative to losing this knowledge is capturing it in structured form before the expert leaves - not as free-text documentation, but as validated, attributed judgment items with the expert's name on each one. Situation, standard procedure, expert practice, deviation, deviation reasoning, conditions, confidence. Nine fields that preserve not just what the expert does, but why they do it, and under what conditions the standard procedure is insufficient.
This is not a documentation project. Documentation projects produce documents that nobody reads. This is a capture-and-retrieval system: the expert speaks, the system structures, the expert approves, and the frontline can query it one-handed on a phone at three in the morning. With a name on every answer. With a date the expert validated it. With the reasoning intact.
The retirement wave is not slowing down. The question is not whether you will lose your senior experts. The question is whether their judgment will still be in the building when they are not.