Performance problems rarely have the cause you think they do.

Documentation errors had spiked after a new EHR implementation. Nursing leadership had proposed a full refresher training cycle for all inpatient staff.

Diagnosis showed the errors clustered around two specific patterns: shift change windows when nurses were documenting under time pressure, and a medication reconciliation screen that required navigating three tabs to enter a value nurses needed dozens of times per shift. The errors weren't random user errors. They were predictable consequences of workflow and screen design.

The fix was a screen modification submitted to the EHR vendor, an adjusted shift handoff protocol that gave nurses a brief documentation-catchup window, and a one-page job aid. No refresher training was delivered.

Internal safety audits at a food manufacturing plant flagged repeated lockout/tagout violations during minor equipment interventions. Two near-miss incidents in six months. Corporate EHS proposed mandatory refresher training for all production staff and supervisors.

A walkthrough with line supervisors and operators showed lockout stations were centralized in two locations, while many routine interventions happened on lines five to ten minutes away. Supervisors under output pressure were tacitly approving skipping lockout for quick tasks. Operators could recite the procedure verbatim. They simply weren't using it for short interventions.

The fix was six satellite lockout stations near high-frequency intervention points, revised supervisor accountability with weekly observation rounds, and shift output targets adjusted for proper lockout time. A short toolbox talk explained the changes. No formal training.

Member satisfaction scores had been declining at a mid-sized health insurance company for three quarters. Leadership had received enough escalated complaints about rude or rushed reps that the VP of Operations had budgeted around $180,000 for an empathy and active listening curriculum.

Diagnosis revealed reps were evaluated on a scorecard where average handle time carried the most weight. Quality monitoring existed but accounted for a smaller share of the composite score. Team leads pulled handle time reports daily and coached anyone trending above target. Listening to calls confirmed reps knew how to be warm. They were rationally choosing not to be, because taking ninety seconds to acknowledge a frustrated member would hurt their numbers more than a complaint would.

The fix was rebalancing the scorecard so customer satisfaction and quality together outweighed handle time, changing the daily report team leads pulled, and a short coaching brief for managers. No rep training.