I've just spent a great day at the Engage Employee Summit. A room full of smart people, sharing smart ideas, but all up against the same wall: we've inherited a definition of engagement that counts the wrong things. Likes. Opens. Completion rates. Attendance.
None of it is anyone's fault. It's just what the tools measure, so it's what we report. It tells you attention happened. It doesn't tell you anything changed.
The numbers say the gap is wider than anyone wants to admit. Gallup's State of the Global Workplace 2026 puts global engagement at 20%, the lowest since 2020 and the first time it's fallen two years running. The bill comes to around $10 trillion, roughly 9% of global GDP. The steepest drop is in the layer meant to hold it all together: manager engagement has fallen nine points since 2022, to just 22%.
Read that back. Only one in five people at work are genuinely in it. The rest are present but not really there. Doing the work, attending the town hall, nodding along at the strategy, and emotionally investing about as much as they would in an airline safety demonstration.
Gallup is measuring the workplace, not the feed. But it's the same instinct in both. People have learned to spend their attention carefully, and they spend it carefully everywhere now, including at work, where the all-hands and the strategy deck are just more things asking for attention nobody agreed to give.
So the real question isn't why workplace engagement is falling. It's why an entire industry built to fix it keeps making it worse.
We're measuring our own exhaust fumes
Here's the trap. We built tools that count likes, opens and completion rates because likes, opens and completion rates are countable. Then we decided they were the goal. They aren't. They're the post-rationalisation, a way of feeling good about feeding the same machine that caused the problem in the first place.
A viral post that changes nobody's behaviour is just attention you rented for an afternoon. It's a receipt. Proof that something happened, not proof that anything changed.
Real engagement has to be defined before you chase it, and it's different for every piece of work. The honest measure isn't "how many people saw this?" It's "what did we want people to do, think or feel differently ... and did they?"
Make an ethics film, and engagement isn't the view count. It's a measurable rise in people calling the whistleblowing line because they finally trust it'll be heard. Run a safety campaign, and it's fewer incidents on the floor six months later. Define the outcome first. Then the work has somewhere to go.
And the forces working against you just got worse
We've spent fifteen years building an economy that runs on attention. Tristan Harris, who co-founded the Center for Humane Technology, calls it a "race to the bottom of the brain stem": platforms competing to hook the oldest, most impulsive part of you, the part that reacts before you decide to.
It worked. And people are tired of it. There's a quiet, growing pushback; a sense that our attention is being farmed and that we'd like it back. That instinct doesn't stay at home. It bleeds into work, where the all-hands is just another thing in the queue.
Then AI made content effectively free. Anyone can produce something polished in seconds. But polish isn't depth, and people feel the difference even when they can't name it. The result is an economy drowning in surface. More content, less of it worth a moment of anyone's time.
So you've got attention that's harder to earn than ever, a measurement system that doesn't notice when nothing changes, and a content economy producing more noise by the day. That's the recession.
What actually holds
Deloitte had to put mandatory ethics training in front of 470,000 people across 150 countries, the kind of brief that usually ends in a completion-rate spreadsheet and nothing else. Instead of another module, they asked for something different. So we made Dilemma, a drama series people chose to watch, talked about, and asked for the next season (It's now in season four). Compliance became culture.
Nothing about that result shows up on the dashboards most of the industry would have used. Completion rates, sure, but completion rates were never the point. The point was half a million people who think differently about the choices in front of them. That's an engagement number you can't put on a dashboard, and it's the only one worth chasing.
There's a reason it holds. Nick Shackleton-Jones calls it affective context: we don't store information, we store our emotional reaction to it, and reconstruct the rest from there. You'll forget a hundred ordinary commutes and remember the one where you nearly crashed. Strategy that stays abstract gets nodded at and ignored. Strategy told as a story people recognise gets acted on, because it's their team, their choice, their stakes.
The 20% number Gallup found isn't an information problem. It's a story problem.
If you've got something that matters and a room that isn't listening, that's a conversation worth having.
