There is an entire industry built around employee experience. Surveys, platforms, awards, conferences, certifications. Companies invest real money in it, sincerely, year after year.
And the scores barely move. Engagement numbers across the developed world have been flat or declining for a decade, through every wave of new programs. Which raises a question that deserves a serious answer: why does so much EX investment buy so little experience?
My answer: because of an assignment error. We diagnosed the right problem and handed it to the wrong department.
What the experience actually is
Strip away the vocabulary and ask what genuinely shapes a person's experience of work. Not the values on the wall. Not the summer party. The texture of the ordinary week:
How long they wait for decisions that block their work. Whether their tools cooperate or fight them. Whether anyone can tell them who owns what - and whether two people give the same answer. Whether finished work stays finished, or comes back for a third round because the requirements were never really decided. How many meetings stand between them and the job they were hired to do. Whether asking an uncomfortable question surfaces a gap or earns a reputation.
Every item on that list is a property of the company's operating system - its decision rights, its structure, its processes, its workload design. The employee doesn't experience the culture deck. The employee experiences the operating model, daily, at point-blank range.
Employee experience, defined honestly, is the sum of operational friction a person meets per week/month/year.
The assignment error
Now follow what happens when the scores drop.
The results go to HR. Always to HR. A department full of capable, well-intentioned people - who own almost none of the levers that manufacture the experience. Decision rights sit with line leadership. Process design sits with operations. Workload sits with planning. Tooling sits with IT and budget owners. Structure sits with the executive team.
So HR does the only rational thing: it acts on the levers it does own. Communication. Recognition programs. Events. Training. Wellbeing initiatives. All of it aimed at the perception layer - how people feel about the friction - because the friction layer itself is out of reach.
This is how a company ends up running mindfulness sessions for a team whose real problem is that every deliverable waits eleven days for one person's approval. The intervention isn't wrong because it's soft. It's wrong because it's aimed at the symptom, by the only department that was given the case but not the keys.
And there's a second-order cost. The team watches the response and draws the obvious conclusion: leadership either doesn't understand the problem or doesn't intend to fix it. That conclusion shows up in next year's scores - which go to HR.
What operations-owned EX looks like
The reassignment is conceptually simple: treat employee experience as an operational quantity, measured and fixed where it's produced.
That means measuring friction, not only sentiment. How long people wait on decisions, by team and decision type. How often finished work returns. How many ownership-ambiguity moments occur in a normal week - the "who actually decides this?" events. What share of a role's hours goes to producing versus coordinating. How long it takes a new hire to reach clarity about who owns what.
None of this requires monitoring individuals. All of it describes the system people work inside - and as I argued in The Survey Illusion, sentiment data still has a seat at the table: it's the thermometer that tells you something hurts. The friction data tells you what to operate on.
Then the fixes go to the owners of the levers. A decision bottleneck is a COO agenda item. A rework loop belongs to whoever owns that process. Ownership gaps get closed by the executive team, in writing. Employee experience stops being a program with a budget and becomes a property of how the company is run - reviewed where operating performance is reviewed, because it is operating performance, photographed from the employee's side.
Why this matters more than it sounds
Because retention follows friction more often than we admit. People rarely resign over a missing perk. They resign over the Tuesday - the eleventh week of waiting for decisions, redoing finished work, and absorbing ambiguity that someone above them declined to resolve. The exit interview will say "better opportunity." The opportunity was partly just a company with less friction.
And because the cost asymmetry is absurd. Perception-layer programs cost real money and move scores marginally. Friction-layer fixes - naming a decision owner, killing a double-approval, writing down what was decided - often cost nothing and move both the scores and the output. When the same intervention improves how work feels and how much work ships, you've stopped trading employee experience against performance and discovered they were the same variable all along.
What HR actually owns
None of this argues that HR is dispensable - the opposite. HR owns things operations never will: talent systems, development, fairness, the craft of understanding people, and the instruments that detect that something hurts. What HR should stop being asked to do is repair machinery it has no wrench for.
The honest division of labor: HR holds the thermometer and the people expertise. Operations owns the furnace. EX improves when both sides know which is which.
Employee experience is the employee-facing surface of your operating model. If the experience is bad, the operating model is bad - and no program aimed at feelings will fix what the structure keeps producing every Monday morning.
Fix the week. The feelings follow.