Recognition Is Not a Result: The alarm was never the problem. What happens after it rings is.

At 3am on a hospital ward, a monitor beeps. A nurse walks past it without looking up.

Not because she doesn't hear it. Because she's heard it correctly, four hundred times already this shift, and on three hundred and ninety of those occasions the alarm was accurate and irrelevant — a patient shifting position, a sensor slipping half a centimetre. The alarm did its job. It kept being right about something that didn't matter enough to stop for. So she has, entirely reasonably, learned not to stop.

Hospitals have a name for this. Alarm fatigue. The Joint Commission designated it a National Patient Safety Goal in 2014, after enough missed critical alerts made the pattern impossible to treat as anecdotal. The finding: a signal loses its power to produce action exactly in proportion to how often it correctly repeats itself without anything following from it. The alarm doesn't fail by being wrong. It fails by being right, over and over, at no cost to anyone who hears it.

I hear a version of that same silence in almost every leadership team I sit with.

The wrong question

Most leadership teams, watching their own version of that monitor — a strategy that keeps drifting, a number that keeps disappointing — are asking: once we've clearly seen the problem, why hasn't it started fixing itself?

That question assumes recognition and correction are two ends of the same motion — that seeing clearly is most of the work, and acting on it follows almost automatically. The right question is different: what, specifically, is built to turn this alarm into a response instead of background noise?

Most organisations have never built that. They've built very good alarms.

What the data confirms at the top

Protiviti and the University of Oxford surveyed 852 C-suite executives globally this June. Not whether AI transformation was underway, resourced, or sponsored — whether the executive personally believed it was driving revenue growth. 30% said yes.

Put that beside McKinsey's tenth annual B2B Pulse, published the same month: only 4% of B2B companies report genuine confidence in a differentiated value proposition — despite 91% expecting to hit their growth targets this year.

Neither number is a knowledge problem. These are the people with the most visibility in the building. Seventy percent privately doubt the thing they're funding. Ninety-one percent expect an outcome that four percent believe they're positioned to deliver.

One CEO put it to me plainly, a few weeks after her own AI review had gone the way these reviews usually go: “I already knew. I’ve known for two quarters. I don’t know why knowing didn’t change anything.” She wasn’t describing a gap in her own understanding. She was describing a gap in her organisation’s Adoption System — the machinery, or the absence of it, that turns a leadership team’s private conviction into something the rest of the organisation actually does differently. Nobody had ever built that machinery. They had simply gotten very good, quarter after quarter, at naming the problem out loud.

Why seeing it clearly was never going to be enough

There's a term for the assumption underneath this: the Information Deficit Model, first named in 1980s science communication research, describing the belief that scepticism is caused by missing facts — and that supplying them changes behaviour. It's since been substantially discredited for one clean reason: people don't act on information the way the model assumes. What actually predicts action is closer to what a 1943 Iowa study found when tracking farmers deciding whether to plant a proven, higher-yield seed corn. The yield data was available to everyone from day one. Almost nobody planted on the strength of it. What predicted adoption was whether a farmer had watched a neighbour's field outperform his own for a season first. The information was universal. The motion was local — one field, then the next.

Recognise, worth noting, comes from the Latin recognoscere — to know again. A backward-facing word by construction. It confirms something you already suspected. It was never built to point forward.

What this means for your organisation

None of this is a leadership failing, in the boardroom or on the ward. A nurse who's learned to walk past an accurate alarm is responding rationally to four hundred repetitions with no differentiated signal. A C-suite that's recognised a gap and watched nothing move is responding to the same pattern, at a slower cadence, over quarters instead of shifts.

I've sat on both sides of this — inside the room where the number was right and nothing moved, and later, outside it, watching from a vantage point where the missing piece was obvious. The hospital fix was never “make clinicians care more.” It was redesigning which alarms sound, and building a protocol that turns a correct alert into a required action rather than familiar noise. In my experience, the organisational fix is the same shape: not a louder alarm, but a structure that makes recognition travel like a neighbour's field, not a data point everyone already has and nobody has acted on.

Recognition tells you the alarm is accurate. It doesn't tell you what to build next.

If your organisation has already had its version of this alarm, and you want help building what comes after it — a structural diagnostic here https://the-friction-point-diagnostic.netlify.app/ takes eight minutes and tells you exactly where to start