Metrics Part 1: The Cardinal Rules
Updated: Feb 27
By Hillel Glazer, TCV Growth Partner -
Metrics aren’t trivial and there’s no discussion about performance without them. Likewise, there’s no discussion about operations without performance. So just like the transitive property of mathematics (A=B, B=C, then A=C), there’s no discussion about operations without metrics. Since we already agree that metrics are critical, one would think that all business leaders know—
- why they’re needed,
- how to identify the right metrics,
- how to get them,
- how to use them, and
- what to do with them.
One would think that, but clearly, given how rarely metrics do well at what they’re needed to do, how often metrics are just lip service, and how regularly metrics go terribly awry, many of us are disappointed by the reality. It’s no wonder that wiring up an operation for metrics is rarely done well.
Sure, metrics need to be tied to objectives and all that yadda yadda. And let’s say it’s actually safe to assume that metrics, like goals, need to be “SMART”. (Please don’t tell anyone you had to look that up.) And yes, when we have explicit programs like Six Sigma and whatnot we obtain and crank on all manner of metrics. So, what’s left? Isn’t deriving metrics from goals and defining them SMARTly enough?
Apparently not. One glance at the Wells Fargo debacle brought to light in the 2016-ish timeframe disarms any argument that merely tying metrics to strategy is as far as metrics need to be defined. (Want a great big pit in your stomach? Read Don't Let Metrics Undermine Your Business in the September-October 2019 issue of Harvard Business Review by Michael Harris and Bill Tayler.)
One assumption underlying many leaders’ deployment of metrics is that they have completely worked out the underlying operation from which the metrics are collected. Anyone reading my posts knows how farcical that assumption is. But let’s not even go there (yet). Let’s start with five cardinal rules for metrics. Why start here? Because they’re CARDINAL. Break these rules and nothing else matters. In fact, go find a failed metric. I’ll wait.
You can just about be guaranteed that it broke one of these rules.
Although my work focuses on operations, I also am of the belief that operations touches and holds everything else together. Therefore, the following rules are not just about operations. They’re universal. Getting metrics right matters in any and every element of business. Let’s list the rules then comment a bit about what happens when each is violated and why each one deserves a spot on the list.
1a. A metric must never EVER about people. (Hence, never to be used punitively.) 1b. A metric must be easy to get and unobtrusive. 1c. A metric must be useful to those whose work produced the metrics taken.
2. Know which decision(s) the metric is tied to.
3. Know which actions drive that metric (and the decision).
4. Which way is “goodness”?
5. See #1.
You want data about the work being done, not the people doing it. Violating rule #1a in any way at all creates a toxic environment which, besides being awful, will cause the metric to be gamed and therefore you’ll get garbage data. Violate rule #1b and the data will likely be stale at best or disruptive to the very work you’re trying to measure—and therefore skewing that same data—at worst. Meanwhile, if rule #1c isn’t in place you won’t have willing partners among your people to help make changes for the improvement of the outcomes. Besides not knowing if or when changes are needed, they wouldn’t be able to be proactive about sniffing out overheating wires before they bloom into fully involved building fires. (Side note: Giving your people the autonomy to make changes isn’t about the metrics but is certainly a part of leveraging the benefits of metrics.)
A rule #2 violation results in pointless data. The ultimate purpose of any metric is to feed into a decision. Knowing a metric “just to know it” gets dangerously close to a rule #1 violation anyway. But even if there were no rule #1, there plenty of other more productive rabbit holes to go down into than those created by metrics that aren’t anchored in decisions.
Many times there are factors and circumstances that contribute to results that are not in anyone’s control. This doesn’t mean that nothing can be done about these factors and circumstances. However, it does mean that there’s no basis for trying to influence the work producing the outcomes. In other words, violating rule #3 is like pushing a string uphill. If you don’t have control over what’s influencing the work, any metric from it has no basis. This is true whether the influence on the metric is truly out of your hands, or you merely don’t have a handle on the process.
Though it may seem trivial, and having “SMART” metrics ought to avoid this trap, it can be easier said than done to know which direction is “good”. Is more of something good, or is less of it better? While closely tied to rules 2 and 3, rule #4 merits its own place because unless you’re literally guessing—which clearly has its drawbacks—before you invest the time and energy in deploying a metric, you want to know which way is the right direction you want to see the metric move. Only with this understanding can you then begin to place any sort of expectation on what it will take to make it move.
Finally, let’s say you’ve worked your way though the first four rules and you’re ready to declare a metric ready for issue. Take precautions to double-check that along the way you haven’t lost sight of rule #1. With all the work that it may have cost to get through the rules, by the time you’ve negotiated your way to metric victory, the metric may have inadvertently devolved back into something that will violate rule #1 and there goes whatever trust you may have built up to get there.
Tying metrics to strategy is only ever going to work if they don’t violate these rules. There is no hidden formula, trick of the trade, or secret sauce to getting metrics right. They must adhere to the rules above. It does help, however, to have a sounding board. The best sounding boards include your own people—specifically the people whose work the metric reflects, as well as someone entirely objective, perhaps to facilitate.
In part two I’ll get into properly framing metrics-based decisions and also flat-out hand you three metrics every business on the planet can start with. Whether hunkering down for a strategy session or trying to find the path to better performance, if you don’t want to wait until Part 2 drops, drop me or one of my partners a message to get working on metrics that will get you there. Hillel@TCV-Growth.Partners