• Thanks for stopping by. Logging in to a registered account will remove all generic ads. Please reach out with any questions or concerns.

Metrics Gone Wild: The Military’s Dangerous Obsession

daftandbarmy

Army.ca Dinosaur
Reaction score
25,003
Points
1,160
You know an article is a good one when you can just feel the bitterness of first hand experience in the sub-text ;)


Metrics Gone Wild: The Military’s Dangerous Obsession​



“That job is more than 30 days old. What’s the story there?” asked the colonel, his eyes peering over his reading glasses.

I have no doubt that my face contorted the way it tends to when you hear something that sparks confusion and disbelief at that same time. Confusion, because he knew full well why the job was more than 30 days old, a shortage of specialized manpower due to summer rotations and a gap in repair parts availability driven by a lapsed production contract. Disbelief, because we’d had this same discussion days earlier. But this time he was throwing me under the bus in front of the one-star Assistant Division Commander for Support chairing the division maintenance meeting.

Better to subject a subordinate to a public bloodletting than explain why a particular metric had not been achieved.

WHEN METRICS GO WRONG

Metrics by themselves aren’t a bad thing. In goal-oriented organizations, they provide measures with which to gauge progress toward achieving those goals. Briefing the status of maintenance “jobs” over 30 days old, for example, ensured that leaders remained focused on equipment readiness, and that logistics expertise could be leveraged when systemic issues arose. However, when those leaders failed to maintain a proper focus on readiness or neglected to address known systemic issues – such as critical personnel shortfalls – it wasn’t unusual for them to simply lie or clear the way for the nearest bus to run down a subordinate in a crowded conference room.

“You really took one for the team today,” he later remarked as we left the conference center. Yeah, let’s just ignore the tire tracks on my back.

Choose the right metrics and use them correctly, and success tends to follow. The right metrics lead to informed assessments that not only help to achieve success, but also reveal patterns that can foreshadow systemic problems. When metrics go wrong, however, the results are often brutal. You don’t just fail to achieve designated goals; you often end up so far afield that you can’t remember why you set those goals in the first place. Generally, there are three ways metrics go wrong: they are misused, miscalculated, or misunderstood.

In one of my graduate classes, students complete a case analysis in which misused metrics are at play. In that case, students are confronted with a real-world situation where employees at a factory are rewarded with time off for accumulating a certain number of accident-free days. It was a perverse incentive: rather than improve safety, the metric led to a reluctance to report on-the-job injuries. The misuse of the metric was only revealed after an employee who had been struck by a forklift tried to “walk off” a severely broken leg.

Miscalculated metrics can be just as bad. In the example of the 30-day-old maintenance job, the colonel’s staff had miscalculated the metric and insisted that I was wrong when I initially raised concerns. It wasn’t until they discovered their mistake that panic began to settle in. Faced with an avoidable human error, the colonel had a choice to make: accept the blame for his staff’s mistake or let someone else take the fall. We all know how that turned out.


 
Too many organisations value that which they can measure, rather than spending the time to figure out how to measure what they value....
Or even what has value. There are some things that are worthless still being bandied about....
 
Yes but pie charts..........

You need data to present a colourful and confusing power point showing that you indeed are worthy of your paycheque.

I once calculated that a certain pie chart had cost the Department $49,000 in man hours to produce, they stared blankly at me as they can't see that all data collection has a cost of it's own.
 
I have just started working with a non-Federal agency, and can already see the bafflegab....insight generation....KPIs...reach objectives...OKR methodology.....all the things I also saw at DND.....
 
I had a boss who wanted everything documented for stats. It was arduous and, frankly, meaningless as our role was more reactionary than having a consistent workflow to monitor.

I remember making him feel like an ass when he asked me what my stats were for the week and replied "I got about 21 minutes of ice time in the 3 games we played, but I had 3 assists and 2 goals...."

The uproarious laughter from the shop was worth the extras...
 
Ah yes, "Everyone needs metrics!" Added several hours of work every week only to disclose... that when my team got information on time, we turned it around on time, but when it arrived late and we were already past the time we owed it to the next level, we didn't make the timeline.

New metric time! On every document we produce, add a checkbox for very senior leaders to tick off "This information was useful" and space for comments, and send us that feedback. After a year of data gathering, and 100+ documents submitted, a grand total of one response.
 
Ah yes, "Everyone needs metrics!" Added several hours of work every week only to disclose... that when my team got information on time, we turned it around on time, but when it arrived late and we were already past the time we owed it to the next level, we didn't make the timeline.

New metric time! On every document we produce, add a checkbox for very senior leaders to tick off "This information was useful" and space for comments, and send us that feedback. After a year of data gathering, and 100+ documents submitted, a grand total of one response.

I had a happy client whose key success metric was 'I don't get angry phone calls anymore'.

Job done ;)
 
The key to useful metrics is good operational analysis. The first step in good operational analysis is actually doing some of it.
 
Back
Top