When Standards Fail Quietly

You get what you inspect, not what you expect.

Louis V. Gerstner Jr., former CEO of IBM

What the supervisor didn't see

We were mid-way through a Gemba walk on linen handling — a routine check that had become urgent after a patient left a child's thermometer inside a bedsheet, which then went through the laundry and destroyed the equipment. We'd decided every piece of linen needed to be inspected before washing, without exception, and we wanted to see the process for ourselves.

The supervisor demonstrated first: running each sheet and pillowcase through a loop formed by his thumb and index finger, an O-ring pass that would catch anything left behind. Clean technique, clearly explained. He then asked a staff member to repeat it.

She got it wrong. Instead of the O-ring, she simply shook each piece out with both hands — a check that would miss exactly the kind of object that had caused the incident in the first place. I didn't correct her. Instead, I turned to the supervisor and asked whether the technique had just been demonstrated properly.

He said yes.

The standard was never in question. The verification was.

That single answer told me more about how standards actually hold in an organization than the sheet-folding technique itself. The staff member's mistake was ordinary — new technique, imperfect first attempt, entirely fixable with a correction. The supervisor's mistake was the one that mattered: he watched the wrong version happen in front of him, and confirmed it as right. Not out of dishonesty — more likely, he'd already decided in his head that the standard was being met, and stopped actually watching for the gap between what he expected to see and what was in front of him.

This is where most standards quietly fail. Not at the point of teaching, and not even really at the point of doing — but at the point of verification, when the person responsible for catching a deviation has stopped looking closely enough to catch it. Over time, what begins as an exception can become accepted practice, not because anyone deliberately lowered the standard, but because no one noticed the gap often enough to correct it. And because not every linen staff member was even present that day, "100% compliance" wasn't something we could assume from one good demonstration — it was something that had to be checked, person by person, every time, with someone whose attention hadn't gone soft.

Redesigning the standard, not just re-teaching it

A second Gemba walk, a different department, made the opposite point. Patient feedback had flagged something inconsistent about room orientation — people not knowing where the slippers were kept, not sure how to reach the nurse call button, small things that added up to a sense that no one had properly walked them through their own room. We asked the nursing team to demonstrate what they normally covered.

It was brief: nurse call button, bathroom, television. Nothing else.

The gap here wasn't laziness or a bad verification moment — it was that the standard itself had never been properly designed. Room orientation had been treated as one task, owned by one team, when it was actually two: a clinical orientation and a non-clinical one. Nursing could cover the nurse call button and clinical safety aspects well enough, but meals, wifi, the bedside phone, concierge services and room amenities sat outside their natural scope. We eventually split the process into two parts, handed the non-clinical half to the concierge team, checked it against every concierge staff member, and verified it again on a follow-up Gemba walk. The specific complaints we'd been getting simply stopped.

Two Gemba walks, two different failures. One was a verification problem — the standard existed, but no one was checking closely enough. The other was a design problem — the standard as written couldn't actually be delivered by the people holding it. But both only surfaced because someone senior enough asked a second question instead of stopping at the first. "Was the technique demonstrated properly?" is a different, harder question than "do we have a technique?" — and so is "walk me through exactly what you cover, step by step" compared to "do we do room orientation?" Both first questions get an easy, self-confirming yes. Both second questions force someone to show their work rather than describe it, and that's where the real answer showed up.

Which raises the less comfortable version of the same point: who's asking senior leaders that same second question?

Repeat it. Demand it. Live it.

Underneath both Gemba stories is a pattern I've found myself returning to repeatedly. At the frontline, the challenge is learning and applying the standard consistently. At the supervisory level, the challenge becomes demanding it from others — and noticing when it isn't actually being met, even when everyone, including you, has already decided it is. The linen example failed at that level. The room-orientation example eventually succeeded because someone looked closely enough to question whether the standard itself made sense in the first place.

At the senior level, there's no one left to run that same second question on you. The standard stops being something you verify in others and becomes something you either live or don't, in full view. And I'd like to say our own leadership team passed that test cleanly, but the truer story is closer to the opposite. For over a year, we ran meetings that started thirty minutes late as a matter of routine, with no real explanation offered or expected. "Are we disciplined about time?" would have gotten an easy yes from any of us. It took someone finally asking the harder version — walk me through what actually happens at 9am, not what we say happens — this time pointed at ourselves, before anything changed. We committed, during a January workplan retreat, to start and end on time regardless of who wasn't in the room yet, and to stop treating lateness as the default anyone could quietly opt into.

It's a less flattering example than I'd like it to be, but that's rather the point. The supervisor's costless "yes" and a year of senior leaders tolerating their own lateness are the same failure at different altitudes — both are what happens when the person best positioned to enforce a standard finds it easier not to look too closely, including, and especially, at themselves.

Where standards actually break

None of these failures — a staff member's technique, a supervisor's inattention, a standard nobody had designed properly, or a leadership team's own tardiness — happened because anyone lacked the standard itself. Everyone involved could have told you, if you'd asked, what "good" was supposed to look like. What failed each time was the check one level up. Someone with the standing to look closely either didn't see the gap, or didn't recognize it as significant enough to address. The only real difference between the supervisor and us was that someone senior eventually checked him. For a year, no one checked us.

Standards rarely collapse in a single decision. They erode through tolerated exceptions. A supervisor overlooks a shortcut. A manager accepts an incomplete handover. A leader allows a meeting to start late because everyone is still arriving. Each decision seems inconsequential in isolation. Taken together, they redefine the standard far more powerfully than any policy document ever could.

I still think about that supervisor's "yes." It would have cost him nothing to say what he'd actually seen. It cost us more than a year to say the same thing about ourselves.