The Engagement Survey Nobody Acts On
A 2,000-employee services firm runs an annual engagement survey with a 78% response rate. Scores have been flat-to-declining for three years; 'I would recommend this as a great place to work' fell from 71% to 64%. HR produces a 40-page report with 60 metrics sliced by department. Managers report they don't know what to do with it, and only 3 of 22 departments set any follow-up actions last year. Leadership is skeptical of spending on another survey and asks HR to prove the survey drives anything. HR has one analyst and a modest budget.
1. Diagnose why the survey program isn't producing change, and reframe what HR should measure to earn leadership's continued investment.
Read the worked answer
The program is failing at the translation and action stage, not the data-collection stage. A 78% response rate is strong; the 40-page, 60-metric report is the problem-it gives managers data without a decision. With only 3 of 22 departments acting, the survey is producing measurement without consequence, which erodes both manager trust and employee trust (employees answer, nothing changes, scores fall).
The reframe: leadership doesn't need more engagement scores; it needs a line from a small number of drivers to business outcomes. HR should identify the 3–5 survey items most strongly associated with the outcomes leadership cares about (turnover, client satisfaction, absenteeism) using existing data, and show that relationship. That converts the survey from a scorecard into a diagnostic and prioritization tool.
HR should also stop defending the survey as an end in itself and instead propose a leaner instrument tied to action. The proof leadership wants is not 'the survey is valid' but 'acting on these items moves these numbers.'
2. Design a lean operating model-one analyst, modest budget-that turns survey results into department-level action and lets you demonstrate impact.
Read the worked answer
Simplify the instrument. Cut to a focused set of items (a short engagement core plus a few driver questions), freeing analyst time from reporting to analysis.
Give managers a decision, not a report. Replace the 40-page deck with a one-page department scorecard: 3 priority items, the department's trend, and a menu of evidence-based actions per item. Managers choose 1–2 actions and commit to them.
Make action the metric. Track action-plan completion by department (target: raise from 3/22 to most departments) and priority-item movement next cycle-not the composite score alone.
Close the loop with employees. Require managers to report back 'you said / we did,' which is the cheapest lever to rebuild trust and response quality.
Demonstrate impact with existing data. Have the analyst link priority-item movement to turnover or client metrics in the departments that acted versus those that didn't. Even a simple before/after comparison gives leadership the causal story it asked for, within one analyst's capacity.