Measuring performance is an essential part of management, but there are some right ways and wrong ways to think about metrics. When done right, metrics are essential inputs to decision-making, driving new insight, and shaping decisions. Done wrong, metrics can become a waste of energy—and create a sheen of progress that distracts you from addressing real challenges.
Here’s how to avoid common pitfalls when you think about how to develop effective performance metrics.
1. Before you pick metrics, think about the toughest managerial challenges you’re facing. What data would lead you to make different, better decisions?
Many managers believe their metrics should be a fixed set of high-level measurements to monitor performance in each area they manage, rather than as a tactical aid for the most significant decisions they’re facing. Seize the opportunity to define metrics that will help you see things you otherwise wouldn’t—and that will not just warn you of a problem, but give you new insight about how to fix it. That may mean measuring quite narrow things, measuring more in certain areas than others, and changing your metrics as your challenges evolve.
2. Don’t disqualify non-quantitative indicators.
Often, managers believe that metrics have to be quantitative, focusing particularly on data that lends itself to nice, time-series charts. But many things aren’t actually meaningfully measured that way. Don’t discount the power of non-quantitative and even subjective measures—sometimes they can be the most powerful and direct indicators of how you’re actually performing. If you’re implementing a new plan, sometimes the best metric is simply “did we meet our most recent deadline or not?” That’s a purely binary measure. Subjective data—such as customer feedback—is often much more impactful when presented as a selection of raw quotes as opposed to some sort of synthesized score. Of course, representative data needs to be gathered thoughtfully to minimize bias, but in practice, most quantitative measures aren’t immune to bias, either. The goal is to have indicators that help you to make different (better) decisions that you otherwise would, and if the most useful indicator isn’t a number, so be it.
3. Commit to setting targets, trajectories and triggers.
Many times, I’ve heard executives say, “We want performance metrics, but we’re not ready to set targets.” Having a metric without a target is a bit like flying with an altimeter but not knowing how high the mountains are. If you can’t contextualize a value in terms of how good or how bad it is, you’ll almost certainly not be able to use that metric to support decisions and choices, which is the fundamental purpose of having metrics. In addition to end-point targets, spend some time up front discussing the trajectory you’d like to see. Discussing trajectories often uncovers important issues regarding how change will actually be driven. A similarly valuable exercise is discussing what management actions you’d trigger if a metric reaches a certain value. These kinds of conversations can really help to make the value and purpose of performance metrics clear and real.
4. Be willing to measure new things.
Many organizations expend enormous resources gathering data, and there’s often pressure to pick metrics that use available data. If you’re trying to implement a new strategy or achieve different results, you often won’t get there measuring things in the same way you’ve measured them in the past. Remember that the seven most-limiting words in organizations are “We don’t cut the data that way.” Be resolved to measure things in the ways that give you the clearest, unequivocal insight, not muddy, “could be” messages.
5. Don’t try to use data that isn’t trusted.
If, every time you show a particular metric, you enter a debate about whether the data is valid, look for a different metric. I’ve been in countless meetings where more time is spent discussing all the caveats about why a particular measure may be flawed than discussing what it implies if it’s accurate. A measurement that nobody trusts isn’t going to drive any meaningful action. If your team responds to the question, “What would we do if this metric dropped by 10%?” with an imperious Captain-of-the-Titanic style, “Of course we wouldn’t do anything, because clearly, the data would be wrong” … the metric isn’t serving its purpose.
6. Keep it simple.
Tracking lots of metrics and having a dashboard that looks like the cockpit of a 747 doesn’t necessarily mean that you’re managing more effectively. If you’re spending most of your meetings just getting through the data rather than engaging in high-quality problem-solving, you have too many metrics. Ensure that you’ve got a process where you can decide to stop measuring things that aren’t proving useful. A handful of metrics is sufficient for most organizations if you’re selecting metrics that are truly essential for guiding key initiatives forward. And if you find you need to dig deeper on a particular issue, you can always do some extra data gathering and analysis where it’s needed, rather than regularly collecting and displaying mountains of data that aren’t essential to current management decisions.
This post first appeared at Portola Advisors.
Tim Gallagher helps companies create strategies and build operational capabilities. A former Partner with McKinsey & Company, he has worked in a wide range of industries, including airlines, business services, electronics, consumer products, and manufacturing.