The Seven Sins of Performance Measurement

Not long ago, I asked a client about his company's metrics for performance. Excited, he reached for a binder that was four inches thick. As my eyes widened, he told me that the following week he would have another binder for the month that just ended. Yes, the binder in his hand held the metrics for just one month.

The episode perfectly highlights how performance metrics—the units of measurement a company uses to track and improve operating performance—are a lot like laws: We keep adding fresh ones without getting rid of any (or getting much in return). I mention it now because in my previous column, I noted that BP's evident carelessness about safety standards can be explained, in part, by poor metrics. Which raises an obvious question: What are the right metrics?

All too often, organizations set measures based on financial targets. That's not a bad thing because they represent what the business needs to accomplish. What is frequently missing are metrics that are aligned with what the customer wants. Finding the balance between the two is the challenge.

So when clients come to me to ask what the right metrics are I usually answer: Make sure there are metrics in place that represent the needs of both the customer and the business as a whole, focusing on the end results you want to achieve. Avoid the seven deadly sins of metric performance:

1. Vanity: This is the sin of using metrics that you have mastered, or that make you look good, rather than help drive improved performance results. Most organizations, for example, measure on-time shipping because that is what their IT systems can report most accurately. It's tougher to capture when something is delivered to the customer, but that's the metric that is far more meaningful to … the customer.

2. Provincialism: This happens when you let organizational or department boundaries such as budgets determine your metrics. One of our clients rewarded its employees in the order-entry department based on order accuracy. However, the order-entry folks were taking so much time double- and triple-checking orders that the company was missing its scheduled delivery dates. Guess who didn't get their bonuses? The delivery department, because it was measured on delivery—which was delayed by the very activity the order department was rewarded for.

3. Narcissism: This is when a company measures performance from its point of view, rather than the customer's perspective. For instance, if a computer software company were to ship 9 of 10 components, it could give itself a 90 percent score. Not bad, except the customer couldn't do anything without all 10 components. It would rate the company's shipping performance at 0 percent.

4. Laziness: This is the mistake of simply assuming what's important to measure—or measuring what's easy. For instance, an electric power utility assumed that customers cared about speed of installation, only to find that customers cared more about the reliability of the service. Doing a modicum of homework about your customers is the first step in discovering the right metrics.

5. Pettiness: Companies too often focus on a small part of what matters, rather than on the totality, or end results. In the case of BP (BP), for example, we learned that it was focused on reducing the number of OSHA-recordable accidents, probably because it feared government retribution. What it ignored was ensuring that the oil rig equipment was working properly.

6. Inanity: Many companies create metrics without any consideration for the consequences. Take the example of a fast-food chain that chose to measure waste, based on how many cooked chickens went unsold at the end of the day. So managers responded by not cooking any chicken until orders were placed, thereby turning fast food into slow food.

7. Frivolity: Then there are companies that just aren't serious about metrics at all. This doesn't necessarily mean having no—or too few—metrics. The company that generated the four-inch binders was guilty of this sin. Any company that doesn't take the business of measuring performance seriously won't have anything to measure after a while.

Having the right metrics in place makes all the difference. Take an auto insurance company we had as a client. Instead of charging clients of a certain age group the same rate, as its competitors were doing, this company realized it had a lot of behavioral data, which it was able to use to determine risk. Subdividing age groups based on risk allowed it to charge different rates for members of the same age group. The insurance company certainly had higher analysis costs than its competitors, but the company also had far better margins.

The beauty lay in the fact that the company attracted low-risk clients with better rates and priced high-risk clients to its competitors. This might sound like a simple idea, but it took the right metrics to get it done.

Before it's here, it's on the Bloomberg Terminal.