Turkey Shoot – Jack Welch on Underperforming Employees

Chairman Jack, or ex-Chairman Jack, is lecturing at my alma mater, and his class, Conversations with Jack Welch, has gotten a fair amount of play in the press of late. One of his rants, errrrr, conversations is about underperforming employees, and he’s quoted as saying “If you’ve got 16 employees, at least two are turkeys.” Later on in the widely-circulated article I read, there’s a further reference to “not bothering to improve the performance of underachievers in the bottom 10 percent of the company’s work force,” which Jack has an answer for. Show them the door. (I like this shot of Jack: it looks like he's getting ready to strangle on of those turkeys.)

I’ll look past the little arithmetic discrepancy here: 16 employees/2 turkeys = 12.5 percent, not 10 percent. (Noticing things like that is just the Sloan School MIT quant jock in me.) More to the point, I don’t know what he bases this little piece of wisdom on, but like a lot of “fast facts” it makes sense on the surface, but doesn’t really stand completely up to scrutiny.

It’s the difference between absolute and relative standards. In absolute terms, you indeed have all employees who are performing up to snuff. No, this is not a case of Lake Woebegone-ism, in which all employees are “above average.” But, especially when there is a small sample/small company, it is entirely possible – practically and statistically – to get decent performance out of everybody. The greater the number of employees, of course, the greater the – practical and statistical – likelihood that you will, indeed, have some turkeys. But 10 percent or 12.5 percent strikes me as arbitrary.

Speaking in relative terms, however, by definition there will always be a bottom 10 percent. You may have to keep looking for reasons to put someone in the bottom 10 percent, but you will certainly be able to find them.

But routinely and definitionally declaring that the bottom 10 percent must go doesn’t make any sense to me. It sounds okay on paper – and who can argue with wanting to raise the bar. But doing an annual or quarterly purge of the bottom x percent strikes me as a recipe for all kinds of bad behavior on the part of managers and employees, and bad morale, as the witching hour for determining who’s on the bottom of the pile this round comes up. It strikes me that this can work – sort of – when there’s an objectively measurable way to determine who’s on last: sales results, production quotas. But where the standards are less quantifiable, watch out.

This is not to say that managers shouldn’t get rid of their “turkeys.” As a manager, you know it when an employee is just not going to work out. Second chance, yes. Third chance, maybe – if the circumstance are really extenuating – but there does sometime come the point where you really need to part company with an employee. And guess what? As often as not, the employee “in question” knows that they’re not a fit. They may not want to be unemployed, but they may be relieved to go.

I had one “problem child” whose performance was incredibly erratic. But she played me like a harp, sensing perfectly when I was about to lower the boom and employing self-preservation tactics that coupled improved performance (to the extent that I would question whether there really was a problem) and a personal sob story. After I got suckered a few times, I sat her down and we worked through an exit strategy. Turns out, she’d knew things weren’t working out and she’d been looking for a job. Did she turn out to be someone else’s problem child or Golden Girl? Not my problem.

Jack may not be right about the percentages, but sometimes you have to call a turkey a turkey and fire away.