If this is you Your dashboard looks great after an automation rollout, and you've cut the staff who used to do that work.

A Great Metric Can Mean Someone Is Quietly Fixing It

The dashboard shows 94 percent of patients leaving with a follow-up booked. One scheduler spends every afternoon undoing what the software booked.

This essay takes on: Why do I get more work every time I do a good job?

A framework essay: the terms are defined as they arrive.

The argument in 7 lines
If this is you
Your dashboard looks great after an automation rollout, and you've cut the staff who used to do that work.
The claim
A good metric can be held up by someone repairing the system's failures off the record, and her competence removes the evidence that she's needed.
The mechanism
Software marks bookings complete, a scheduler fixes them afterward off the record, the dashboard counts her work as automation, and leadership reassigns the other schedulers.
Who pays
The scheduler, whose unlogged afternoons hold up the number, and patients who show up for appointments that don't exist when she's out.
Who can correct it
Most institutions can't see the problem to correct it, because nothing was built to tell planned exception handling from repair hidden by the completion count.
What would change it
Ask what happened before when these people took a week off and what happens now, and watch for performance rising while tolerance for absence collapses.
The objection it answers
Any measure becomes a target. The cheapest ways to pass the absence test, staffing up, logging the repairs or fixing the software, all improve the operation.
Reading settings

A gauge. It pushes on the person to go faster. Lanes sort them, and they go down the low one.

A scheduler spends her afternoon undoing what the software booked that morning. She is a composite of people and incidents from different organizations.1 It sends follow-ups to clinics that closed, to times the patient already said they couldn’t make, to specialists who stopped taking that insurance two years ago. She calls, rebooks, apologizes, and catches most of them.

Upstairs the dashboard reports 94% of discharged patients leaving with a follow-up booked, up from 71% back when three schedulers worked the queue by hand. Two of those three have since been reassigned. Every one of those numbers is accurate, and the appointments in the record exist, though most of them were booked twice.

Her afternoon appears nowhere. There’s no field for it.

The metric she is repairing is the metric that says she is not needed.

Diagram
The concealment feedback loop
Conscientious repair keeps dashboard metrics green, leading leadership to cut staffing under the illusion of automated efficiency, destroying system tolerance for absence. How to read this
View source
%% title: The concealment feedback loop
%% caption: Conscientious repair keeps dashboard metrics green, leading leadership to cut staffing under the illusion of automated efficiency, destroying system tolerance for absence.
flowchart TD
  classDef actor stroke-width:1.6px;
  classDef system stroke-width:1.2px;
  classDef failure stroke-width:2.2px,stroke-dasharray:6 3;
  classDef repair stroke-width:2.4px,font-weight:bold;

  S["Software books appointments<br/>(High silent failure rate)"]:::system --> R([Scheduler catches errors]):::actor
  R -- "unlogged repair labor" --> D["Dashboard reports 94% follow-up<br/>(Metrics green)"]:::repair
  D --> L["Leadership inference:<br/>'Automation succeeded, schedulers redundant'"]:::system
  L --> C["Staffing cuts / Reassignment<br/>2 of 3 schedulers removed"]:::failure
  C --> V["Tolerance for absence collapses<br/>Fragility concealed until sick leave"]:::failure
  V -. "scheduler out with flu" .-> B([Downstream clinic collapse]):::failure

Her competence removes the evidence

We know how to read a bad number. A queue backing up, a complaint rate climbing, an error log filling. Those are legible, and institutions have machinery for them.

We’re much worse at reading a good number that someone is holding up.

She is not hiding anything. Nobody asked her to conceal the rebooking; nobody would have stopped her from mentioning it. She doesn’t mention it because there is no distinct thing to mention. It is the job. The job has always included fixing what arrives broken, and the fact that what arrives broken now arrives from software instead of a fax machine does not change how it feels to fix it.

A survey asking about “unlogged corrective work” will probably miss this. She’ll say things are busy without separating repairs from the rest of her day. Asking her to walk through yesterday’s bookings gives the question something concrete to attach to.

And her competence is what removes the evidence. If she were worse at her job, the follow-up rate would fall and someone would look. Because she is good at it, the number holds, and the software’s failure rate is stored nowhere except in her afternoon.

Where the hidden work shows up

What happened the last time she was out?

Before the software, when a scheduler took leave, the queue moved slower. Bookings dropped a few points and recovered the following week, so the operation bent without breaking.

After the software, the week she had flu, the follow-up rate stayed at 94%, and the clinic spent the next month absorbing patients who showed up for appointments that didn’t exist. The number held because the number was never measuring the thing. What collapsed was downstream and delayed, so nothing connected it back.

Before this system, what happened when these people took a week off? What happens now?

Not “does it degrade.” Everything staffed by humans degrades when the humans are away. Watch for divergence: reported performance improving while tolerance for absence collapses. A system that genuinely reduced work should get more robust to a missing person. When those two lines separate, the gap between them is the work that isn’t being counted.2

Staffing and operational records probably already hold both halves of the comparison. Cross-referencing them against the software’s introduction gives an investigation somewhere to start. It does not isolate the software’s effect. Reassignment, demand and the patient mix can all change what happens during an absence.

When the software routes something to a human because it cannot handle it, that is the system working as designed. Exceptions exist. Someone should catch them. In the intended design, the machine takes the volume and the person takes the hard part.

When the software marks something complete and a human fixes it afterward, off the record, that is different in kind. The dashboard already counted it. She is manufacturing the number the dashboard reports as automation.

What separates planned exception handling from repair concealed by the completion count is how much of her work lands on items the system already reported as done.

Most institutions can’t answer that, and the reason is duller than concealment. Nothing was ever built to tell the two apart.

Counting the scheduler

Any measure becomes a target and this one is no exception. But it has an unusual property: the cheapest ways to cheat it are the fix.

An operator who wants to pass the absence test can staff up, so a missing person is survivable. Or find the informal repairs and make them logged, resourced duties, which puts the labor on the books and ends the false efficiency. Or fix the software so there is less to repair.

Each of those improves the operation. One serious evasion remains, push the absorption somewhere the telemetry cannot see, onto contractors, onto the patients themselves as a second phone call, and the metric will drive exactly that if the rest is closed off.

An institution that cannot say who its systems depend on knows only how well those systems appear to work. That is a different fact, and one that people are producing on its behalf without being asked and without being counted.

The scheduler is not a failure of the software. She is part of it. Two of the three schedulers were reassigned on the strength of a number she is holding up, and the saving was booked to the software. Whether the institution knows that changes nothing about who collected it.

Notes

1

The scheduler is a composite. The double-booking, the reassignment and the flu week each come from a different organization; no single one is a case I can name. Read her as the shape the pattern takes. The term for what she has become is not mine. See Lineage for where it comes from.

2

Organized labor has known this for over a century. Work-to-rule, following the written procedure exactly and withholding the informal judgment that normally fills its gaps, is better read as a demonstration that the specification was never sufficient, and it works because the gap it exposes was always there.

Run the check this essay describes

The remainder, and who flexes

Rules compress reality, and compression guarantees a residue that someone absorbs. Rank is the right not to be that someone, and the dashboard never learns their name.

A clinic template, sixteen visits as they actually ran, and a choice of who absorbs the minutes the rule can't hold. The utilization tile reads 100% in every state.

Open the check

When they get you wrong or count on you

12
Who pays
The scheduler, whose unlogged afternoons hold up the number, and patients who show up for appointments that don't exist when she's out.
Who can change it
Most institutions can't see the problem to correct it, because nothing was built to tell planned exception handling from repair hidden by the completion count.

Continue reading

What to read next

Continue with the next essay in Conditional authority, or open where this essay sits: its pattern, instrument and toolkit.

Related essays

New essays by email

One situation a week, the essays that explain it, and one question to put to the institution.

New essays by email

One situation a week, the essays that explain it, and one question to put to the institution.