HomeBusiness Ethics

Microsoft sacks stack because it lacks

Like Tweet Pin it Share Share Email

There was a time it would have pained me to give Microsoft a compliment. But no more.

There are those in the industry who see all of Microsoft through a Windows 8 lens. Not that it has anything to do with this week’s missive; my opinion is that Windows 8 is best understood as Vista 2 — what matters isn’t Windows 8 itself, but what Microsoft learns from Windows 8 as it figures out Windows 9.

No, this isn’t a Windows 8 column. It also isn’t a Visual Studio column, although from what I’ve read, Visual Studio is outstanding. Nor is it an Azure column, even though Azure might just be the most interesting and enterprise-ready PaaS platform available, especially given the enterprise need to integrate the development environment (that would be Visual Studio) and identity management (Active Directory) into the whole shebang.

Nope. This isn’t a technology column. It’s a human-resources-management column (aka “human capital management,” aka “how you treat the men and women who work for you” column).

Microsoft has made a cutting-edge discovery: Stack rankings are a terrible idea and must be stopped.

To understand why stacked ranking … insisting that ten or twenty percent of every manager’s direct reports must be classified as “doesn’t meet expectations” … is a terrible idea you have to start by understanding where and in what circumstances they were a good idea, namely, General Electric when Jack Welch first took the helm.

Welch had been put in charge of a flabby, complacent company. He was quite sure it was overstaffed, and equally sure it had more than its share of mediocre-or-worse employees. It wasn’t hard to go from there to an inescapable inference: GE also had more than its share of managers who were quite comfortable drifting into the future on the efforts of mediocre employees.

These managers were too numerous to terminate en masse, so the only practical alternative was to give them no choice in the matter. Every year, every manager had to cut the bottom ten percent of their workforce.

Which had two salutary effects. First, General Electric got rid of its least profitable employees. And second, managers learned it was okay to terminate employees who weren’t worth keeping.

But just because it made sense in that restricted set of circumstances doesn’t mean it makes sense as an HR practice that pervades American industry.

As pointed out here years ago (“Staff inflections, KJR, 9/11/2006):

 Jack Welch popularized cutting the bottom ten percent of the workforce every year. Because it was Jack Welch, too many business leaders decided it must be a great idea. If you inherit a complacent, flabby organization this might be just the ticket … for a year or two. Beyond that limit, forget it — it’s both statistically and socially invalid.

Statistically: If you trim the worst performers, and choose strong replacements, then after two years your average performers must be well ahead of the industry average. And if you give other managers credit for brains, those they’ve terminated were the ones who didn’t do their jobs well, dragging down the average among the currently unemployed. If you can continue to strengthen your workforce by churning those rated lowest among your employees, you’ve been doing a poor job of recruiting and retaining great employees. You won’t fix this by continuing to cut.

Socially: How do you think your best employees will respond to the annual ritual? My guess — they’ll find companies that take a more surgical approach to dealing with problem performers. So in the effort to cut your worst employees you also lose your best.

Even worse: The more you churn your workforce, the less employees will trust each other, or you, because trust takes time to develop. Without trust there is no teamwork; without teamwork few organizations can achieve anything important.

Sorry to quote myself so extensively and self-congratulatorily, but I have nothing new to say on the topic, other than to congratulate Steve Ballmer.

Yes, Steve, your name is on several seriously bone-headed moves. Fair’s fair. If you’re going to get nailed for them, you also deserve credit: for making Microsoft’s products far more enterprise-ready than when you took over; and for building Microsoft’s enterprise infrastructure product portfolio. And especially now, you deserve credit for getting rid of stacked ranking at Microsoft.

It’s a fine legacy. Even more important, it just might make it okay for the HR departments in other companies to abandon the practice in favor of alternatives that have the advantages of making … what’s the word I’m looking for? Oh, yeah, now I have it.

Sense.

Comments (4)

  • Wasn’t stack ranking and particularly letting the bottom 10–15% go one of the innovations brought to Enron by Jeffrey Skilling? And, if memory serves, it was one of the reasons for its demise: it lead to a culture of gaming the numbers so severe senior leadership didn’t know they’d hit an iceberg even when the ship was halfway sunk.

    My employer does the version that I suspect most bureaucratic companies do; the bottom group doesn’t necessarily get sacked, but there are only a few high rankings available. If you’re in a group with someone else who’s always gotten the one A, you have very little incentive to try to do more than B work.

  • You may be giving Ballmer too much credit. This move may have been mandated by the new CEO, whoever he may turn out to be.

    You also didn’t mention some other downsides to this system:
    1. Top performers will avoid working together, so each doesn’t get hurt by the other.
    2. People will sabotage top or near-top performers.
    3. Potentially top performers will underperform, so they won’t be sabotaged.

Comments are closed.