6 Reasons Why Your Performance Metrics Aren’t Working

Jonathan Feldman

I am a fan of business performance metrics the way that I’m a fan of maps and compasses: Too much focus on the map, and you forget to do things like see the sights. To stop and meditate along your hike. To greet others on the trail. To climb the tower and see if you still want to be going in that direction, despite all indicators showing you that you are still going towards the destination. Too little, and you might know where you’re going, but you have no idea where you are.

Metrics are the map used by healthy organizations. And, of course metrics are needed, along with the data that drive these metrics. But, still, metrics tend to either be worshipped (“it’s ALL about the metrics!” “Only the metrics matter!”) or get a bad rap. (“They’re not working!” “Why are we collecting these metrics? We never do anything with them?” “It’s just busywork!”)

With all respect: these things are said because, frankly, you’re doing it wrong. Otherwise you wouldn’t hear these things.

I can’t help the metric worshipers. That’s a mental condition. But for those who’ve been burned by metrics, I offer some reasons why they haven’t worked for you.

1. You’re not asking the right questions.

WHY do we collect performance metrics? It’s because we want to know whether something that we’ve done or that we are doing is effective or not. If it’s not, we want to take action to fix it.

Within the department that I lead, we ask four critical questions:

  1. What are we trying to accomplish?
  2. What are we doing about it?
  3. How do we measure if these actions are effective?
  4. Can we measure often enough?

One of my colleagues distills this into “what questions are we trying to answer with the metric?” and that’s not a bad way to look at it, either.

There are a lot of much more complicated ways to look at performance metrics. The most popular one is “SMART” – “Specific, Measurable, Actionable, Realistic, and Timebound,” but I find that most employee’s eyes glaze over when you start using acronyms like this, instead of just asking them in plain language what the heck they want to accomplish.

Try it, you’ll like it.

2. You think that the only answer is “Paradise or Armageddon.”

It is so tempting to say things like “passed network audit!” or “project completed!”, but these things are useless as performance metrics. I call it the “paradise/armageddon” syndrome. To be wonkish, these types of measures are binary quantities; they are yes/no, not shades of gray.

Performance metrics are about acting on the metric, so it is useless to have binary metrics. By the time the metric changes, it is too late to act.

Since the entry into Armageddon is not desirable, do some thinking about what your early warning signs are for Armageddon.

I vastly prefer knowing something like, “how many times have we self audited?” to “did we pass the audit without a major finding?” Of course we want to pass the audit without a major finding. But that’s Paradise or Armageddon. By the time we find out, it’s too late. Checking our progress on self audits is a far better way to make sure that work is being done towards the Paradise goal.

For some activities, the only thing you think you can do is represent progress as “Percentage of milestones achieved,” but be aware that many milestones tend to be “heaven/hell” type of binary quantities as well. And for the love of Pete, nobody really listens to your “60% complete” report-out without rolling their eyes. Seriously.

3. You are not collecting the right data…

I continue to be amazed when people approach IT saying “let’s do some performance metrics.” “It will be data-driven decision making!” Well, where’s your data? “Oh. We thought you took care of that.” No, actually IT does not do that in a vacuum. The business partner understands what the goal is and what data could be available — and what type of new data collection might be possible. IT doesn’t.

4. … Or, you are collecting data but you are not making that data available.

Various business units tend to believe that only they really need their “own” data. That’s almost never true. Maybe there’s a financial interest in engineering data. Maybe there’s a facilities interest in IT data. And so on. But belief generates practices: the practice of creating data silos and not sharing that data with the rest of the organization.

Also, too much organizational data is locked away in desktop spreadsheets. Fortunately, cloud software “spreadsheets” are actually useful data sources, unlike their earlier desktop counterparts.

5. You forget to change metrics when the business objectives change.

Remember the map? Metrics are the map & compass that lead to your destination. But, they don’t define whether the destination is desirable. Is it raining? Is it time to seek shelter? That’s a business decision, not something you consult a map for. Scores of decisions like this trail decision happen every day in organizations. We must seek new maps when we seek new destinations.

6. You have too many dang metrics.

Metrics are like goals. If everything is a top priority, nothing is a top priority. Metrics can be a lot of work and take a lot of time. I’m not saying don’t collect data. But if your strategic dashboard has 52 metrics, you’re doing it wrong. Most successful executives follow a rule of a small number of goals in order to ensure the focus needed to get those goals executed upon. It’s the same with metrics. Sure, you have to measure to manage. But are you really going to actively manage 52 things at once? No. It’s confusing and takes focus away from what really matters.

Originally published August 3, 2017