Pity the employees at a Starbucks in midtown NYC. In a misguided attempt to improve quality, the management posts monthly scores on a variety of metrics. . . without understanding anything about effective use of metrics.
In this photo, you can see the percentage of customers who responded “Strongly Agree” on questions related to store cleanliness, order accuracy, food taste, connection, “going above and beyond,” and other areas of concern.
Measurement is a good idea, but only if it’s done well. These measurements? Not so much.
If you read Mark Graban’s blog or book, you’ll immediately see problems with this chart. For one thing, three data points don’t make a trend. With no upper and lower control limits, the movement in scores is nothing more than management by emoji — we have no way of knowing whether the movement is just random noise in a stable system, or a real signal indicating something significant happened. And why are they looking at the scores monthly? By the time they see a decline, it’s far too late to figure out what the root cause was and how to address it.
Imagine the monthly meetings with team members, where management demands to know why the “Connection” score dropped from 32 to 29 between November and December. How could anyone answer that question intelligently?
At best, reviewing the scores is a waste of time, forcing people to “write fiction”—concoct a justification for something that’s almost certainly random and unknowable. At worst, it’s employee abuse, given that they might very well be evaluated (rewarded or penalized) for these scores.
Starbucks: if you’re going to measure performance, then do it properly. Don’t waste management’s or associates’ time with pointless exercises that doesn’t help anyone to improve.
That’s just torture.