Eleven Retirements, Undocumented Knowledge, No Requisitions
A specialty manufacturer will lose 11 of its 34 maintenance technicians to retirement within 18 months. Their troubleshooting knowledge is mostly undocumented. Two of the departing technicians have refused to be recorded on video. Production runs 24/5, and pulling a technician off the line for four hours costs about $3,800 in overtime backfill. The plant manager's plan is to hire replacements early and pair each with a retiree for two weeks. HR says requisitions will not open until the roles are vacant. You have a modest budget and the plant manager's tentative support.
1. How do you assess and prioritize the knowledge at risk?
Read the worked answer
Not all 11 retirements carry equal risk, and treating them equally will exhaust the budget before the critical knowledge is captured.
Prioritize by frequency × production impact × difficulty to learn:
- Use existing data: downtime logs, scrap rates, maintenance tickets, and safety incidents by equipment and by technician. Look for the failures that only two or three people can diagnose.
- Run critical-incident interviews with each departing technician: "Tell me about the last time this line stopped and you were the one who figured it out." That surfaces tacit diagnosis faster than asking them to document processes.
- Separate procedure from judgment. Written procedures likely already exist; the risk is the diagnostic reasoning that selects among them.
The two technicians who refuse video are not a dead end. Respect the refusal and use structured interviews, decision logs, and having them review a draft guide written by someone else. The test of capture is whether a successor can act on it, not whether it was filmed. Draft one guide with a successor and test it on a real fault before scaling.
2. Evaluate the plant manager's plan of early hires paired with retirees for two weeks.
Read the worked answer
The direction is right; the two-week pairing will not transfer tacit troubleshooting.
- Two weeks is observation, not practice. Watching an expert diagnose a fault builds familiarity, not the ability to do it under pressure. Skill acquisition needs repeated attempts with feedback and increasing difficulty.
- Pairing without structure defaults to shadowing. Specify tasks, a progression, and sign-off criteria: which faults the successor must diagnose independently before the retiree leaves.
- Timing matters. If the pairing happens months before the retiree departs, the successor will be alone when the hard cases appear. Stagger it so pairing continues past the hire date and into the first real failures.
- Use the overtime math to justify the cost. At $3,800 per four-hour pull, a handful of structured sessions is cheap against a line-down event.
Keep the early-hire idea but redesign the pairing: one retiree with one or two successors over several months, task-based, with artifacts captured during real work rather than in extra sessions. Add a retiree on-call or part-time arrangement as a bridge.
3. HR will not open requisitions until vacancies exist. How do you move this?
Read the worked answer
Reframe from a policy argument to a risk-and-cost argument, and give HR an exception path that does not require them to break a rule.
- Quantify the exposure. Downtime cost per hour, overtime backfill, and the probability that a critical fault occurs before a replacement is trained. Compare that with the cost of opening two requisitions early.
- Do not ask for all 11. Ask for the two or three roles holding the highest-risk knowledge. A narrow, evidence-based request is far easier to approve than a headcount exception.
- Engage the plant manager as co-signer. HR is being asked to absorb headcount risk; the plant manager owns the production risk. A joint business case carries more weight than a request from talent development.
- Probe the actual constraint. Is it a headcount freeze, a budget cycle, or a policy? Each has a different workaround: internal transfer, a contractor, an apprentice requisition, or a part-time retiree contract that is not a new headcount.
- Offer a pilot with a review date so approval is reversible.
If the exception is denied, document the risk in the program plan and proceed with capturing the highest-risk knowledge first.