The Numbers Aren’t Good

I spent most of the night tossing and turning and generally unable to sleep because I was being haunted by numbers.

Each year the company is forced to set assorted targets to meet. Some of those are understandably financial of course, but to ‘show value’ for the IT department those metrics also extend to things like issue resolution rate, the number of emergencies we have, crap like that. We get hounded each year to make sure the numbers are up to snuff because if we don’t meet them, it will apparently trigger the apocalypse somehow. Whatever.

Much as I’ve railed on and on about the bad precedent that status reports set, these kind of metrics are kind of the ultimate dilution of reality resulting from them. That is to say, they’re taking the nuanced complexities of IT and dumbing it down so that even the most inexperienced layman can look at them and quickly determine that yes, the company is getting its value for money, as demonstrated by the fact that we beat our high score for last year. Good job all, let’s type in our initials.

It’s garbage though, and not solely in the sense that I hate it and therefore I shall insult it. It’s garbage because the stack of tasks put upon the technical types by management isn’t focused on performing our jobs better, it’s focused on gaming the stats so we look better. And needless to say it’s way beyond the point where there’s real benefit to this. We may have a critical issue which I can solve in 20 minutes. But then I have to:

  1. determine if our monitoring system caught it 
  2. fill out a monitoring report so people can know it was caught/was not caught 
  3. fill out the problem report 
  4. document the timeline and the steps taken to resolve the issue
  5. be on hand for questions if my manager didn’t read #4 for the morning problem meeting
  6. report in the weekly problem meeting on whether or not root cause has been determined 
  7. close the incident ticket 
  8. write up a ‘how to fix it’ doc for our knowledgebase for everyone else (arguably the only useful step here)
  9. document the issue again for my status report
  10. fill out the time spent on ‘the issue’ in our time keeping software, which has now ballooned from 20 minutes to 4 or 5 hours.

Hardly value for money, but every step is counted towards some metric or other, so it gives the impression to executive management that it’s all working.

But on top of that mess, the numbers often don’t realistically apply. We had an issue not long ago where the manager of the problem management team was seeing a problem on one of our websites, and since he gets to do such things, he called it a critical issue. Everyone gets on the phone to troubleshoot, and nobody (including the manager) can replicate it. All of the signs point to user error here, and frankly I’m willing to forgive that and not rail on and on about the incompetent fool in my blog. However, it still got called a critical issue and is therefore logged against our metrics. But since we couldn’t determine a ‘root cause’ that was acceptable to said manager, we have a ding on the metrics: apparently user error doesn’t count somehow. To correct this, we’d need to make some BS up as root cause.

I spend hours, weekly, working feverishly at the command of my management to participate in this surreal, elaborate lie to pretend we’re being as efficient as possible, when we could be vastly more efficient if we dropped the pretense.

The real pisser is that this sort of mentality applies to how the employees are doing as well. I’ve mentioned that we have the annual employee survey of course, and talked about how weird it can be. It’s a part of this too: the big statistical report that comes out of this gets tossed up the chain of command, and the managers closer to my level are tasked with doing what they need to to improve numbers for next year. My team’s approval rating of management was in fact historically low this year: our manager will see no direct fallout from that. He’s merely tasked with providing a rough outline of how he intends to improve it and getting the numbers up for next year (as are all managers). It’s been theorized, rather seriously, that he was being a dick this time around to make the numbers dip to make it better next year.

All of the hopes, joys, dreams, fears, goals, difficulties, frustrations, successes, failures, triumphs, and despairs of the organization, conveniently converted to a discrete number that can be instantly compared to last years’ number to determine if we’re doing a good job.

And it’s not like it’s different anywhere else.