Customers Versus KPIs

You've read it before, in this space and elsewhere: Peter Drucker’s famous phrase, “You can’t manage what you can’t measure.” In fact, it was just 11 months ago that I brought it up. It’s the reason for all the gauges on the dashboard of a jet plane, as well as the indicators on your CRM’s dashboard. Key performance indicators—KPIs—are the data we look at to judge whether an operation is going according to plan, and how to adjust things if it isn’t.

In that November 2015 column, I also brought up W. Edwards Deming’s counterpoint that running a company on visible figures alone was a fatal error for businesses. I’m revisiting these opposing ideas this month and hoping that repetition isn’t a fatal error for writers. Fortunately, if romance novelists are any indication, I should be safe.

The problem with KPIs is that they only tell you about what they’re intended to measure. They don’t tell you why, at least not until you get into deeper analytical models. KPIs are intended to be simple, at-a-glance numbers that anybody can point to and say, “That needs to be higher,” possibly with a checklist of next best actions to make those numbers change. The numbers are abstractions of what’s going on with people—your employees and your customers.

The classic case of valuing numbers over actual service is average handle time in a call center. I know I’ve discussed this before on this page, too. Incentivizing customer service agents to get people off the phone as swiftly as possible makes the clock more important than the customer. Not only does this evoke negative feelings, it acts counter to the idea behind wanting fast calls—reducing wait times and costs. Customers who get the bum’s rush will have to call back multiple times, jamming the lines just as surely as longer calls. Satisfaction drops even though the KPI number gets better.

Sometimes sticking to the KPIs beyond the bounds of common sense can do more than harm your bottom line—it can harm people. Have you ever wondered why, when you’re boarding a flight, the crew are in such a rush to close the cabin door and push back from the jetway? It’s because the airline industry measures on-time performance by when the doors close and the plane departs the gate, even by a few inches. It doesn’t matter if the flight is number 19 in line for takeoff and stays on the ground for an hour, it still left on time as far as the company is concerned.

A few years back, video emerged of airline passengers stuck on a plane for hours and denied refreshments and bathroom access while also being prevented from leaving. There was suffering galore, and the outrage led to a change in policy and law. Now the airline must allow accommodation of needs or offloading of passengers after two hours on the ground—which is still pretty bad, especially when you consider that grounded, immobile planes will often have their air conditioning systems shut off, even on hot, sunny days. This recently happened to my associate Rob Hilsen, an executive with Zendesk. He likened the experience to being a baked potato.

While it’s possible to mitigate the misery by purchasing flight annoyance insurance, which I covered in this space exactly two years ago (pure coincidence, I assure you), it misses the point. Elderly or infirm passengers can die from stuff like this, even if they aren’t slathered in butter or sour cream. On-time departure is at odds with customer comfort, and even the hectic ballet of taxiing aircraft should have some space for humanity.

We need our numbers; the KPI is not a shameful thing. All I’m saying is that we need to remember that there are people behind the numbers, and people don’t like to be fitted into a spreadsheet. 

---

Marshall Lager runs Third Idea Consulting, where he manages to measure CRM, customer experience, and other fun stuff. Check in on him at www.3rd-idea.com or www.twitter.com/Lager.