In a couple of jobs I’ve had, I’ve ended up volunteering for work and finding myself getting a somewhat thankless task: collecting departmental metrics. I also got a chance to cowrite a class on training evaluation, which included more metrics. So I’ve picked up things here and there. I recommend metrics as a side project if you have an analytic bent to your writing and are looking to keep busy between major assignments.
Metrics are a common management tool for demonstrating progress toward some goal: X number of days without an accident, percent participation in the annual charity drive, number of people participating in mandatory training, or number of pages developed in the process of developing a large document. Metrics are a way to put numerical truth to claims of accomplishment when qualitative assessements (“How do you think we did?”) aren’t quite thorough enough.
One of the primary mnemonics of metrics is called SMART, an acronym meaning that metrics should be:
Reasonable enough. But there are other things to consider. For example, are your metrics easy to decipher visually? Are you using a pie chart to convey percentages, line graphs to convey changes over time, or multi-level bar graphs to show changes within a group over time? Is your primary message (number of products with defects) being lost in a chart that also includes number of parts delivered and a host of other statistics? Are you, in fact, telling a clear, truthful story with your graphics?
And a cautionary note, lest you or your managers get too carried away with measuring everything. If “the Bart hand” goes up in a meeting, it usually means I’m about to ask, “Is this really necessary?” So there are times when metrics are NOT an effective measure of progress. Situations where you don’t want to measure include:
- Making a count when you only have one item to quantify (say, one rocket or aircraft) over a long period of time. However, that would be a good opportunity to measure progress by timeline or percentage of work done.
- Measuring something that is unlikely to change much, like the size of your building or plant once completed.
- Counting things that you cannot control or that are likely to contribute to low morale (for instance, number of people laid off).
- Measuring problems without a fix in mind. I once counted the number of times my computer crashed. The point wasn’t to be a pain to the boss or the IT guys–the point was to get a replacement computer. I kept a daily count of “Blue Screens of Death” (this was the late 1990s, when such things were more common), and after two weeks of twice-daily crashes, I presented my metrics and my manager put in a request for a new machine.
As with any other communication product, organizational metrics exist to communicate a message and (sometimes) solve a problem. They also provide a diverting side project that gives you an opportunity to experiment with Microsoft Excel or some other spreadsheet software with graphics capability and thereby demonstrate your usefulness as a heroic communicator. After all, if you know when you’ve succeeded, you’ll know it’s time to cheer, right?
Up, up, and away…
This post is quite informative. I have not had to create a system of metrics from scratch, so it would have been helpful to have a brief explanation following each of the elements of the SMART acronym, especially the actionable part. Nonetheless, it is most enlightening.