Fourteen Retirements and a Manual Nobody Reads
You are an organizational development manager at a 4,200-employee manufacturer with two plants. Fourteen of 31 maintenance technicians will be eligible to retire within 24 months; their average tenure is 22 years. Unplanned downtime costs about $18,000 per hour, and last year three failures were traced to procedures that existed only in experienced technicians' heads. A previous knowledge-capture project produced a 600-page manual that supervisors say nobody uses. Production cannot stop for training, the union agreement limits overtime, and only three apprentices are in the pipeline. Plant managers are skeptical of 'HR programs' and want solutions that show up on uptime and safety metrics. Safety incidents in areas staffed by new technicians rose 9 percent. You have $120,000 and 18 months before the first retirement wave. The VP of Operations has asked for a plan that does not add headcount.
1. Why did the previous manual fail, and what approach would you propose instead?
Read the worked answer
The manual failed because it captured knowledge out of context and gave no reason to retrieve it. A 600-page document is written for the author, not the technician at 2 a.m. with a fault code. It has no owner, no update path, and no link to the task.
Propose knowledge in the workflow:
- Identify the critical tasks and failure modes where lost knowledge causes downtime or safety risk.
- Capture short videos, photos, checklists, and decision rules at the machine, co-authored with the technicians who do the work.
- Build retrieval around the task: QR codes on equipment, a searchable mobile library, and a shift-board link.
- Use retirees as mentors and reviewers while they are still available, not as manual writers.
- Establish communities of practice and a named knowledge owner for each critical system.
Change the unit of success from pages produced to procedures used. Track retrieval, mentor hours, and time to competency. Start with two critical systems in one plant rather than a plant-wide capture program.
Engage supervisors in choosing the tasks; their ownership determines whether the material is used or shelved like the manual.
2. How do you prioritize among 14 retirements and 24 months, given that production cannot stop?
Read the worked answer
Use a risk-based portfolio, not a seniority list. For each retiring technician, score the knowledge by criticality (safety, downtime cost, sole-source), likelihood of loss, and time needed to rebuild it. Interview each person for two hours with their supervisor to name the top ten tasks only they can do.
Rank the resulting tasks across both plants. Put the highest-risk items first: equipment with no backup, procedures tied to safety incidents, and systems where the 9 percent incident rise is concentrated. Defer lower-risk knowledge that is documented elsewhere or easily relearned.
Build overlap, not documents. Where possible, arrange part-time or phased retirement and pair each retiree with a named successor for a defined number of shifts. Check union agreement and local rules before promising schedules. If overtime is limited, use planned shadowing during normal shifts.
Pilot in one plant for 90 days to test the capture and mentoring model, then scale. Publish a simple heat map to the VP of Operations showing coverage by critical system. The goal is not to capture everything; it is to protect the knowledge whose loss would stop production or hurt someone.
3. How do you get plant managers to own this, and how will you measure success over 18 months?
Read the worked answer
Put the metrics in their language. Plant managers care about uptime, mean time to repair, safety incidents, and overtime. Co-design the scorecard with them: downtime on critical assets, time to competency for new technicians, safety incidents in new-technician areas, and mentor coverage. Use the $18,000-per-hour downtime figure to make the case, not an HR participation rate.
Give them control. Let them select the critical systems and the mentors, and put a supervisor on the steering group. Fund the first pilot in the plant whose manager is most willing; a visible win is more persuasive than a mandate.
Measure leading and lagging. Leading: critical tasks documented and validated, mentor hours, retrieval counts, apprentice progress. Lagging: downtime, safety, rework. Baseline both before the pilot and review monthly in the operations meeting, not a separate HR forum.
Report honestly. Downtime has many causes; show the trend and the mechanisms. If the pilot does not move leading indicators, change the approach. Eighteen months is enough for coverage and time-to-competency; it may not be enough to prove a safety effect. Say so.