Metrics, Metrics, Everywhere

If you are in any kind of management position, there seem to be a million metrics to track. You get huge pressure to track a number of financial metrics such as EBITDA, gross sales, net profit, and others, there are sales metrics divided up every which way (by territory, by sales rep, etc.), there are productivity metrics such as cost of goods sold, labor hours per unit sold, etc., and on and on. One can metric themselves insane if they try to simply follow all these things, let alone try to lead the organization via them.

But, lest you think I’m not a fan of metrics, please be assured that I think people who don’t track their business in some careful and meaningful ways are playing fast and loose with the fate of the organization. The issue is what to track, how to track it, and what to do with the results. That’s where some careful thought is warranted.

When it comes to the people side of the business, there is no shortage of interesting metrics to track. As I have mentioned in other places, I am a fan of numerical performance review ratings, and using those ratings effectively. And I like to see people paying attention to things like the ratio of the highest paid to lowest paid employee. It helps keep things real (unlike here and here). Another good one is counting the number of layers there are in an organization — how many people from the top to the lowliest janitor. (I like to see no more than 7 or 8… when I was at Microsoft, there were 13). And one of the first things I like to look at when talking to a manager for the first time is how many direct reports they have (5 to 10 is good, fewer leads to organizations that are too deep, more means chaos).

You can’t let the measurement run the organization, that’s your job.

But in all these metrics I have no hard and fast rules. I’ve seen people manage 14 people effectively. I’ve see cases where people earn and deserve huge paychecks. Like most good metrics, you can’t let the number alone tell the whole story. You have to dig deeper, to understand the situation in its entirety. The metric is just the flag. It’s not “Danger, Will Robinson!“, it’s more like “something is up here, maybe you should take a closer look”. You simply can’t let the measurement run the organization, that’s your job.

There are many reasons to be wary of metrics. First and foremost is the problem of measuring things accurately. Even with the simplest of measurements there can be arguments about what a good rule is. Do temporary employees count in a measurement of layers? Do one-time bonuses really count in a compensation ratio? Can I rely on the sales figures from a certain period of time with all the changes due to returns and unsold inventory in the channel? This is the business equivalent of the Heisenberg Uncertainty Principle, where some things just can’t be measured accurately (or perhaps aren’t worth measuring).

There are many reasons to be wary of metrics.

Then there is the problem of the observer effect, where the mere act of simply measuring something changes the thing being measured. In business, this tends to be the real problem. The second you begin to watch a metric and people find out that’s what’s being measured, they react to that measurement. In the best case, people try to improve that metric as you hoped and expected. In the worst case (a cynic would say in every case), people try to “game” the metric and try to trick their way to higher measurements. They “stuff the channel” right before a quarter end. They don’t show costs until the next quarter to improve profitability. Whatever the game, measuring something almost assures that someone will try to figure out how to win.

And of course we all know intuitively that people don’t do well at measurement. They respond badly to being treated like numbers, or simply cogs in a machine. They round or fudge the numbers when they are asked to measure things. And they make simple human errors. So it’s best not to build systems that rely on only metrics to judge organizational health or success.

Use metrics as one view into your world.

Again, that is not to say I don’t think there is a reason to use metrics. I think it’s important for an organization to find the things that they have in shortest supply, or that are the few keys to success, or that make a real difference in the long term health of the business, and measure those. I’ve worked with several companies to build such tools. And they prove extremely useful shorthand ways of finding out how things are going. A business dashboard if you will.

The key to all of this is to remember what metrics are: one measure of some small part of the operation. No one metric can hope to capture the whole organization. No metric is immune to mis-measurement. No metric is impervious to being gamed by those affected. And no metric is worth going insane over. Use metrics like you use all of your senses (sight, smell, hearing, taste, touch, and intuition): as one view into your world. A meaningful blend of of all of these inputs usually provides the best information.