Performance Measurement

Consider the source

In defining metrics, the kind of data you'll need and the sources of data you'll use become important considerations. Ask yourself:

• Where will you get the needed data? In some cases, your organization may already collect data that you can use—such as employee participation in certain training programs. In other cases, you'll need to gather the data yourself, which may require you to set up new processes and systems. Data may also come from external sources.

For example, your company may have hired a vendor to track the number of employees enrolled in benefits programs.

• How will you gather subjective data? Performance metrics can require objective or subjective data.

For example, to track the metric "Percentage increase in sales revenue," you'd gather objective data (changes in sales revenue over a specific period of time).

But how would you collect information to track the metric "Percentage increase in customer satisfaction"? In this case, you'd need to gather subjective data—such as comments or ratings from customer surveys. Even subjective data must be quantified.

For example, expressing customer satisfaction ratings on a scale of 1 to 5—to enable comparison and highlight opportunities for improvement.

• Should you use composite data? If you want a big-picture view of your group's performance, consider using metrics that require composite data—data from numerous sources aggregated into one number. Well-known examples include the S&P 500 stock market index and the J.D. Power index of customer satisfaction with new cars and ratings of new-car quality.

An example of composite data a company might use to measure performance would be a "brand index," which might combine, say, advertising budget, percentage of target audience reached, and brand impression (what people think of the company's brand, measured in a survey). Because the component measures are dissimilar, they would need to be mathematically adjusted to weight them properly.

• Will your data be reliable? While defining metrics, do everything you can to ensure the reliability of the data you'll need to gather.

For example, will the data be up to date? Sufficiently detailed? Accurate? Auditable? Will the data be available frequently enough for you to reliably track performance on your metrics?