All Posts in Metrics

|In: Business Ethics|Last Updated:

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

|In: Career Management|Last Updated:

Capitalism, Winston Churchill might have said but didn’t, is the worst system of economics … except for all the others.

Its virtues are celebrated whenever politicians speak and it has many, not least its ability to balance supply and demand, including demand nobody knew existed until someone invented a gadget to satisfy it.

On the other side of the ledger is its susceptibility to feedback loops, both positive, leading to bubbles, and negative, leading to economic depressions.

I just thought I’d share that cheery thought. Bubbles and depressions have nothing to do with this week’s topic.

What does have a lot to do with this week’s topic … certifications and what to do about them … is the corrupting influence capitalism has on so much of what it touches, the desire for wealth being the root and all that.

Take any truly meaningful certification and you can bet those responsible for maintaining its integrity are either insulated from the economic impact of the certification or benefit from keeping the certification restrictive.

Start with something simple — grades. Once upon a time I taught an IT-related topic in a local university’s graduate program. I awarded A’s to those students who excelled, B’s to those who did well, C’s to those who achieved a basic level of understanding, and D’s to everyone worse.

The Dean asked me to change my grading. Why? Most of the program’s students were eligible for tuition reimbursement from their employers, but only if they maintained a B average. The school’s revenue depended on lax standards.

Now, a related development threatens to make earning a college diploma a deeply meaningless achievement. Increasingly, graduation rates are considered an important measure of a school’s worthiness.

KJR hasn’t delved into the seriously dull subject of metrics for some time, so as a reminder, there are four metrics fallacies: Measuring the right things wrong; measuring the wrong things, right or wrong; failing to measure something important; and extending measures to those with a personal stake in them.

This one’s easy. Once colleges and universities are assessed based on graduation rates, they’ll have three obvious courses of action: Make admissions more selective; educate students better; or A’s for everyone!

Which looks easiest and most certain to you? Yup, and easy-and-certain is a good predictor of future behavior.

The target graduation rate for colleges and universities seems to be around 90%. For contrast, the U.S. Air Force Academy only passes 75% or so, which makes sense once you figure whoever passes might be the one you rely on to shoot down the enemy plane shooting at you.

Then there’s the Bar examination. Rates vary widely by state, from 41% in Louisiana and California to New Mexico’s 85%. The lawyers who make up state Bar Associations benefit from restriction, as everyone who passes increases competition in an already overpopulated field, and everyone who passes without being fully qualified further discredits a field with a poor reputation.

Compare that to the MCSE. Oh, wait, you can’t, because that little statistic doesn’t seem to be available. My guess: Those who teach Microsoft technology benefit from high pass rates. Microsoft benefits by having a large population of IT professionals certified in its technology.

But both rely on a perception that receiving the certification is a difficult achievement (for all I know it is — I’ve never tried to earn one.) Publishing pass rates would let everyone know where test administrators draw the line.

Not to pick on Microsoft or the MCSE — it’s mentioned here because it’s well-known, not because it’s better or worse than any other vendor-managed certification.

Quite a few respondents challenged last week’s contention that certifying bodies should ban the use of certifications as a qualification for job applicants. And I have to admit, that’s probably too extreme: There are plenty of fields where the certifications are downright reassuring, from medicine to the law to construction to beautician (okay, I admit, I don’t actually get that one).

So I’ll back off a bit. Instead, before you use a certification to screen applicants for a position, look carefully at the incentives associated with its administration. A shortcut: Whatever else, certifications administered by independent non-profit associations are more likely to be focused on protecting the integrity of a profession than those administered by for-profit vendors.

And if your plan is to receive a certification, including a college diploma, some related advice: Given a choice between studying for the exam and for gaining actual knowledge and ability, go for the knowledge and ability.

Both will help you get certified. But only one will help you in the job once you get in the door.

Comments