The hidden cost of a departing plant expert

The retirement announcement may come with six months’ notice. That sounds generous until the team tries to list what the departing expert knows.

She knows which machine alarm can wait and which one will ruin the shift. He knows why one customer’s orders need a different sequence. The official procedure says what to do under normal conditions. The expert knows when normal conditions do not apply.

The hidden cost is not the information that leaves. It is the slower decisions, repeated mistakes, and cautious work that follow when nobody else can interpret the situation.

Do not begin with “document everything”

That request is impossible and demoralizing. Experienced people do not carry a neat manual in their heads. Much of their knowledge appears in response to a machine sound, a strange order, or an exception they recognize before anyone else does.

Start by identifying consequential work. Ask supervisors and peers where they seek the expert’s help, what waits when that person is away, and which decisions others hesitate to make. Review recent interruptions, quality holds, schedule changes, and customer escalations.

Choose a few situations where delay or error matters. The goal is not to preserve every preference. It is to make the operation less dependent on one person’s judgment.

Watch real work and ask about the fork

Pair the expert with the person who will inherit the responsibility. Observe a full task or decision cycle. When the expert changes course, ask what signal prompted it.

The useful knowledge often sits at a fork: if the material behaves one way, adjust the setup; if an order has a certain characteristic, confirm the drawing; if a supplier changes a lot, inspect before release. Capture the condition, the options considered, and the reason for the choice.

Examples are more revealing than general interviews. Use recent cases, including one that went badly. Ask what the expert noticed, what source was trusted, who was consulted, and what would have made the decision easier.

Record uncertainty honestly. An expert may use judgment because no reliable rule exists. Turning that judgment into a rigid checklist can make the process worse.

Find the private systems

Critical knowledge often lives in personal files, email folders, notebooks, browser bookmarks, and carefully arranged spreadsheets. Inventory them with the employee’s participation and respect legitimate personal boundaries.

For each operating asset, decide where it belongs, who should have access, and what context another person needs. Move shared records into company-controlled storage. Check whether macros, links, or local applications depend on the expert’s account or computer.

Look beyond files. Vendor technicians may call the expert’s mobile number. A supplier portal may use an individual login. Automated reports may run under a personal account. These are continuity problems as well as access problems.

Build capability, not an archive

A folder of videos and procedures can create false comfort. The successor has to perform the work.

Use a simple progression: observe, perform with guidance, perform while the expert observes, then handle an exception independently. Let the successor explain the decision afterward. This reveals where instructions are incomplete or confidence is missing.

Test runbooks by asking a capable employee who did not write them to follow them. If that person cannot find an input, recognize a stopping point, or verify the result, revise the runbook.

Some skills need more than one backup because they affect every shift or site. Others can rely on a vendor or documented escalation. Match the investment to the consequence and frequency of the work.

Improve the process while preserving the knowledge

Knowledge transfer can expose controls that should not be preserved. Perhaps the expert routinely corrects bad master data, bypasses an unreliable integration, or approves work that should have another owner.

Separate three outcomes. Preserve sound judgment that remains necessary. Fix avoidable process defects. Retire practices that depend on outdated conditions or inappropriate access.

Involve the expert in these choices without making that person solely responsible for redesign. The operating leader owns the future process.

Technology can help where it fits. Searchable work instructions, recorded demonstrations, maintenance histories, and decision logs make knowledge easier to find. An AI search tool may help employees retrieve approved material, but it cannot recover context that was never captured or guarantee that old instructions are correct.

Plan the departure mechanics

The final weeks should not be a scramble for passwords. Confirm system owners, transfer vendor contacts, reassign scheduled reports, update escalation paths, and remove unnecessary individual dependencies.

Set a cutoff for new work that only the departing expert can perform. Otherwise, urgent tasks will continue crowding out transfer until the last day.

After departure, watch operating signals connected to the role. Are decisions waiting longer? Are scrap, downtime, expedites, or support calls clustering around certain work? Review those signals with the successor and adjust the support plan. Do not interpret every early question as failure. Responsible successors ask questions.

Start before there is a notice

The best time to reduce single-person dependency is during ordinary operations. Cross-train around the tasks that repeatedly stop when someone is absent. Include continuity in supervisor reviews and system changes. When a new workaround becomes essential, capture why before it becomes folklore.

Leaders should know which five people would be hardest to lose unexpectedly and what work depends on each. That list may be uncomfortable, but it is more actionable than a broad “tribal knowledge” initiative.

EdgePoint can help map operational and technology dependencies when a departure exposes wider risk. The heart of the work remains internal: give another employee enough real practice to make the next difficult decision.