January 9, 2015
“Get Me Out of This Bucket!”
There’s a trap waiting behind traditional employee engagement surveys.
The moment the worker hits “Submit,” she’s destined to be dropped into a bucket. She’ll be stereotyped. Her motivations will be assumed. She may be vilified. If she lands in the wrong bucket, she could be targeted for firing.
All because she answered honestly to the survey, and the people behind the survey or at her company misused the information.
It’s time to stop pouring people into buckets. If the practice continues, more and more employees will lie on the surveys, as they should in those situations, and most employee surveys will become a waste of time and money.
Many consultancies bucket the people who take their surveys. (One actually refers to people in one of its compartments as “hamsters.”) The primary offender, because of its prominence and imprecision, is Gallup. Gallup hit the market hard in the late 1990s with its Q12 survey, making hundreds of millions in revenue and inspiring a number of copycats that picked up some of its bad habits.
Gallup drops employees into three buckets: “engaged,” “not engaged,” and “actively disengaged.” Only three in 10 workers are in the shiny best bucket, we’re told. Roughly half fall into the big bucket of “not engaged.” And about two in 10 are in the contemptible container of “actively disengaged.”
The worst thing about the buckets is what some consultancies say about the people who on the day of the survey gave answers that got them dropped into a middle or left-most bucket. Bucket language got us to a point where we talk about “disengagement” as a character flaw. It led to talking as if there were three kinds of employees, not a wide range of situations, largely external to the worker, which create a wide range in motivation and commitment.
Here’s how Gallup describes its middle bucket:
Not engaged workers can be difficult to spot: They are not hostile or disruptive. They show up and kill time with little or no concern about customers, productivity, profitability, waste, safety, mission and purpose of the teams, or developing customers. They are thinking about lunch or their next break. They are essentially “checked out.”
That description is poorly written. It assigns the blame to the employee, not to the situation in which she finds herself. And, based on in-depth studies my colleagues and I conducted for Widgets, the description is flat-out wrong. People in the middle of the range display high levels of responsibility and maturity.
It gets worse. This is Gallup’s allegation against the two in 10 workers who they classify as “actively disengaged.”
Actively disengaged employees are more or less out to damage their company. They monopolize managers’ time; have more on-the-job accidents; account for more quality defects; contribute to “shrinkage,” as theft is called; are sicker; miss more days; and quit at a higher rate than engaged employees do. Whatever the engaged do — such as solving problems, innovating, and creating new customers — the actively disengaged try to undo.
It is remarkable that any business gets done if 20 percent of the workforce are borderline criminals. But the work gets done because this definition, too, is false.
Of course some people are laggards. Some try to work as little as possible. A few will steal. Some will heave fragile customer packages over fences. We’ve all seen the YouTube videos. But just as the company would rush to mention that misbehaving employee is not representative of their brand, the bulk of the frustrated or demoralized employees, if they could speak up, would rush to emphasize they are absolutely not trying to “undo” what their better-managed colleagues are getting done.
Nobody wants the kind of job that reduces him or her to the point of not giving a damn. No one wants to be on that end of the scale. And for the vast majority of people most frustrated by their jobs, the answer is not to blame them for their so-called “disengagement.”
The bucketing of employees leads to several additional problems for leaders, managers, and the people they supervise.
- It grossly simplifies the range of engagement. Engagement runs along a wide continuum, something of a snowdrift blown to the right. On a given day, an employee is most likely to find herself somewhere in the well-populated middle of the curve, but she could be anywhere. Every parent and high school student knows this. Standardized tests have long reported the percentile score of the student: “Tommy scored at the 79th percentile in math and at the 72nd in reading.” If schools started bucketing kids into just three categories – passing, not passing, and actively failing – there would be pitchforks and torches at the next board of education meeting. Yet, for some reason, we’ve put up with it for employees.
- Based on only 12 five-point-scale questions, Gallup’s curve can only have 49 distinct points. How many are at each point and which scores put you in one bucket or another are closely guarded secrets. It’s therefore impossible for all the journalists and engagement commentators to know the details beneath the generalities Gallup supplies. They have to take Gallup’s word for it. In our age of transparency, telling people just enough to mess with their heads is a disservice to the working public and their employers.
- No one tells the employee which bucket she’s in. Perhaps she can guess, but it would be just that – a guess. If it did tell her where she landed, chances are the survey would not say to her face what some advisors say about her and her bucket buddies behind her back.
- The percentages in each bucket are not really meaningful. When you slice up something close to a bell curve, tiny movements of the knife to the left or right dramatically change the percentages in the two slices. This is Statistics 101. Right next to the 30 percent who are supposedly engaged are another five or 10 percent who are closer to the border between the two groups than they are to the midpoint of their assigned group.
- Bucketing, combined with infrequent measurement, masks the dynamic nature of people’s attitudes about their work. People have good days and bad ones. Some get praised by their managers right before the survey; others have managers who yell at them. People get new bosses. New policies are announced, for good and bad. Their coworkers come and go. Projects dry up or swamp an employee’s personal life. Their engagement goes up and down by small degrees or huge swings in response to the conditions around them. But in the records, until the next survey, people are trapped in the bucket to which they were assigned at survey time.
Described precisely and accurately, there is nothing wrong with bracketing some ranges within the larger statistical distribution of responses employees give to a survey. We do it, as I describe here. But ranges become buckets when a worker is stuck with that label, blamed for being there, and deprived of knowing how she is being categorized.
Of course, no reputable survey company tells anyone at a client organization into which bucket a given employee falls. Her responses are technically “confidential.” But when some team engagement reports show the results for as few as four or five people, it leads to a lot of guessing and, sometimes, hunting down of the suspected “disengaged” as leaders try to ferret out those people who are ostensibly “more or less out to damage the company.” Their time would be better spent figuring out why their employees would grow frustrated or demoralized.
It’s well past time we freed people from the buckets.
Update: Minneapolis-based Modern Survey spammed out the following email on March 12.
About half of all disengaged employees in the U.S. are not looking for work elsewhere. They are comfortable bleeding your organization dry while making other employees do the heavy lifting.
This poses a threat to all employees in an organization. Failing to deal with these employees can cause your top performers to feel resentment and lose their drive to help the organization.
If a friend who was frustrated in his job called me for advice on how he should respond to a Modern-Survey-run employee survey, I would have no other choice than to say, “Mark five to survive.”