KPI

Key Performance Indicators: When your job is so enigmatic, someone else’s job is to devise how to measure it.

To quote kpi.org:

“Key Performance Indicators (KPIs) are the critical (key) quantifiable indicators of progress toward an intended result. KPIs provide a focus for strategic and operational improvement, create an analytical basis for decision making and help focus attention on what matters most.”

That’s a long-winded way of saying: “Measuring how to do your job.” You know it’s bullshit when your concept requires more than one sentence to define.

Personally I think the best way to measure if someone is doing their job is to look at whether or not they’re doing it, i.e. is the job finished. If it is, then you’re doing your job. If it isn’t, then you’re not doing your job. This meta bullshit is entirely unnecessary, so it stands to reason that one day we were given a KPI: Coming up with KPI’s. The amount of KPI’s we could come up with was our KPI. Time to get to work.

We invented all sorts of measures and metrics that could conceivably be reported on, but which did not really matter at all. Still, it felt oddly empowering since when we started putting up these wildly esoteric facts and figures into reports, managers pretended to be completely on board with everything. We were paid to invent metrics that were completely removed from reality and someone else got paid to pretend to understand them.

The jig would be up the moment anyone would ask what the point of all of this was. From a game theoretic perspective though, no one would ever dare question any of it, since everyone would be out of a job when people would realize that absolutely nothing of value is being reported on and absolutely nothing actionable would result from looking at those reports. It’s better for everyone if you just play along.

Sometimes we would go back to our facts and figures when someone who pretended to read our reports would say something like “I don’t like this number, can you check it again?”

So we’d go back to our report, do the obligatory: “SELECT RANDOM hard-coded-number FROM table”, and just like that, we output a number that everyone pretends to be content with, and we give the usual “sorry, we forgot to include x in this, good catch” placation and moved on to prepare the next pointless report.

Every once in a while, an intern or junior analyst might join the team and they’d be anxious about how their reports might be met by managers and stakeholders. I always wanted to tell them that it didn’t matter what they put into those reports, since no one understands them anyway. Also, managers never asked for those reports, we just made them up ourselves.

Occasions when managers might ask something silly like, “can you report on this obscure thing going forward?” were rare, but they did happen. The problem with this is that managers like that don’t last very long, and they tend to move somewhere else after a very short time, and then we’re stuck reporting a single obscure figure that absolutely no one reads anymore. All of the sudden, a single report can slowly balloon and begin to require a higher headcount just to be produced. What used to be a half hour job once a month, can easily become a single team’s entire function for a whole month, which just means a manager needs to be hired to manage a team like that. Then managers like that have to report to their manager or a VP about all the reporting their teams are doing.

All these pointless reports have led to the birth of the analytics software industry. Dashboards that look cool and that users can click around in are being created, but no one looks at them and if they do, they don’t understand them. What could be a spreadsheet of “this is last month and this is this month” is now a dashboard that requires a back-end data engineering team to support and a front-end analytics team to drag and drop stuff around. The sales pitch for this type of software is always the same: You can create a dashboard and your manager just needs a URL and they can go drag and drop and filter whatever they like.

Managers never do this.

Instead, they want you to screenshot your dashboard and put it in a Powerpoint presentation that is full of screenshots of dashboards created by other teams. This is what people mean when they talk about analytics.

And all of this started because teams like ours were given a KPI which was coming up with KPI’s, i.e. bullshit that you can measure.