"We didn't see it coming."
I've heard that sentence from good leaders, standing in the wreckage of a resignation that gutted their team. But they did see it coming. It was in the engagement survey. Six months ago. On page 34.
Twenty years of measuring, and the number went down
Gallup reports that global employee engagement has fallen to around 20%, down from 23% in 2023, and prices the cost of disengagement at roughly US$10 trillion a year — about 9% of global GDP.
Sit with that for a second. We have measured engagement with almost religious diligence for twenty years. The surveys got longer. The dashboards got prettier. The consultants got richer. And the number went down.
If a diagnostic ran for two decades while the patient kept getting sicker, at some point you stop blaming the patient. The problem was never the questions. It's the calendar.
The clock of an annual survey
- March: fieldwork.
- May: results to the board.
- July: action plans cascaded.
- September: any real behaviour change.
Six months between the signal and the steering input. Half a year in which the person you were trying to keep has already updated their CV, taken the call and signed elsewhere.
And the signal was smudged to begin with. Recency bias is well documented: most of us can't reliably recall specific workplace behaviour beyond the last couple of months. So an employee filling in an annual survey isn't really rating their year. They're rating their last few weeks. You are making twelve-month decisions on a few weeks of blurry memory.
Meanwhile, Gallup finds people are 3.6 times more likely to be strongly motivated to do outstanding work when they receive daily feedback from their manager rather than annual feedback.
Mirrors are useful — for what's behind you
Let me be fair to the mirror. A rear-view mirror is not a bad instrument; it's an excellent one. Annual surveys are genuinely useful for benchmarking, board reporting and year-on-year trends. They're simply pointed at a time horizon you can no longer do anything about.
A rear-view mirror tells you what you hit. A windshield tells you what you can still avoid. Most organisations are trying to steer with the mirror.
That's why we built Budaya around frequent, lightweight check-ins: small enough that people actually answer, fast enough that a manager can act on Tuesday on what they learned on Monday.
The old wisdom said: encourage one another daily. Not annually. Daily — because drift is slow and invisible, and frequent contact is the only real defence against it.
One honest caution
This applies to us too. A frequent survey that nobody acts on is worse than an annual one. It teaches people their voice is decorative.
So the shift isn't really annual → daily. It's audit → instrument panel. An audit is done to people, by outsiders, to produce a verdict. An instrument panel is used by the driver, continuously, to make the next decision.
Try this
- Ask two questions every week: "How was your week, 1 to 5?" and "What's one thing that would make next week better?"
- Read every answer within a working day.
- Act on one thing — and tell people what you changed.
The only question that matters here: the last time you learned something important about your team, how old was that information?
Watch the video: Is your culture always looking backwards? · Try the ritual: Friday Two-Question Pulse
