Taking the measure of IT professionals

Like Tweet Pin it Share Share Email

Take everything you think you know about employee compensation. Put it all aside (other than the last two editions of Keep the Joint Running). Read two books: Robert Austin’s Measuring and Managing Performance in Organizations and Alfie Kohn’s Punished by Rewards.

You’ll realize that the two assumptions on which most modern compensation systems are built — that managers can objectively measure employee performance, and that compensation systems should use those measures to create incentives for stronger performance — are pretty much worthless.

I devoted a column to Austin’s book (“Can metrics be saved?” Keep the Joint Running, 11/7/2005). But while I’ve mentioned Kohn in my books, I was surprised to discover I’d never once mentioned him in Keep the Joint Running.

I don’t entirely agree with Kohn. In particular I don’t agree with his rejection of praise as a leadership tool. Still, his core insight — that compensation-as-incentive amounts to bribing employees to perform — is what first drove me to re-think the whole subject.

Credit where it’s due.

We’ve spent two weeks on how to structure compensation. Underneath it all is this chain of logic:

  • Motivating employees is something business leaders are supposed to do. The best you can hope for from your compensation system is to avoid demotivating them.
  • To avoid demotivating employees, compensation systems have to be fair and perceived as fair.
  • To be perceived as fair, compensation systems have to have clear, strong connections to both marketplace value and employee performance.
  • To connect to employee performance, managers must know how to assess it. Not measure it. Assess it.

In business, some employees have jobs whose performance is easy to quantify — factory workers, for example, who process some number of widgets per hour with some (presumably smaller) number of defects.

That some employees have jobs whose performance is easy to quantify has led to the strange conclusion that all of them do. Managers who haven’t fully quantified them just haven’t been clever enough to manage the task.

That isn’t the case. In particular, the list of IT jobs where performance can be quantifiably and objectively measured is quite short. In most shops it’s entirely empty.

Stop trying to objectively measure the performance of IT staff. It isn’t going to happen any time soon, and when it does it will mean we’ve degraded the profession to the status of an assembly line. Assess staff performance instead, recognizing that many dimensions of the job have fuzzy boundaries and subjective qualities.

Here are some ideas for what you might look for:

What each employee accomplished. Keep a list for each employee, which should include the accomplishments of teams to which they belonged. Encourage employees to keep their own lists, too. If you’re leading with maximum effectiveness, you won’t know everything employees achieve.

What they didn’t accomplish. You want self-starters. Employees who are such self-starters that they disregard the work you assign to them, though, shouldn’t be rated as highly as those who get it done. Botched and ignored assignments matter.

How their performance in quantifiable tasks compares to their colleagues. Many responsibilities aren’t quantifiable. Some are. For example: Some developers plow through their share of the maintenance and enhancements queue faster than the rest, with higher levels of end-user satisfaction and fewer bugs. Value this.

The extent to which they pitched in. Sometimes there’s dirt on the floor. You want whoever notices it to grab a broom and dustpan and take care of it. Stuff needs doing. You can’t list it all in advance, put it on job descriptions, partition it all out, and schedule it.

How well they supported their colleagues. “Team” is an overused word. It’s a team when its members align to a common purpose, trust each other, communicate well, and help each other out. You want this. Some employees do more to make it happen than others.

Good judgment. Some employees make better choices than others. The worse ones might do so deliberately through malicious obedience, or because they just can’t figure it out. Regardless of the reason, employees with good judgment are more valuable than the ones who need everything spelled out thoroughly enough that their job is just painting by numbers.

Measuring the performance of IT professionals is an infinitely complex task. Assessing it, on the other hand, is not. You have to take on just three responsibilities.

The first is to know what you value. The second is to communicate it.

The third is to pay attention so you know if you’re getting it.