The engagement survey is necessary. No argument there. An organization should have a way to measure how people feel about work. But the longer you run them, the more you see a blind spot nobody likes to talk about.

The trouble with the engagement survey is that it gets trapped by its own design. Whichever way you build it, you hit a different dead end.

Make it transparent, open for people to say anything, and the results tend to sprawl. So broad you cannot grip them. Everyone complains about something different, and you are left not knowing where to start.

Make it tactical, tied to business numbers with a clear target that the score must go up, and you invite a different problem: people solving it by cutting it short, ticking the good boxes to be done with it. A low score means questions to answer. Once the score becomes a KPI, people manage the score, not the problem.

Here is something seen firsthand. Ask two frontline retail people directly, and you get the same honest answer. Ask through the line manager, and the answer changes. Because if a problem shows up in the results, it lands back on that line manager. The middle has every incentive to make the numbers look good rather than true.

So what actually works

It turns out the thing that proves genuinely useful is not the survey, but an open forum where people talk straight.

There was a dev firm, almost fully remote, that only got together every once in a while, over food and beer. When they did meet, people naturally clustered and talked shop, half casual, beer in hand, standing outside a bar.

What came out of those circles ran deeper and truer than any survey score. Because nobody felt assessed. There was no field to fill that could come back as evidence. Just people talking to people about the work they were actually doing.

Maybe real engagement lives where we do not measure

Real engagement may be driven more by the coffee corner, the breakfast nook, the after-work chat, than by an annual survey. The small things that make people want to show up, talk, and stay are rarely the things that surface in a report.

This is not a case for dropping the survey. It still has its place: a broad view, a trend line, a read on what changed since last year. Keep doing it, it is necessary. But do not mistake a high score for real attachment, and do not overlook that the most effective engagement work might be making space for people to meet and talk straight, with no form in between.

Measure to see the picture. Then build the real thing where the numbers cannot reach.