I knew better.

Last week’s column finished by recommending that you avoid “hiring yourself” when filling management positions. It closed with a clincher: “After all, one of the most basic principles of information technology is to eliminate redundancy.” Jonathon Kass responded perceptively:

“This struck me as an odd statement — probably just because I’ve been focused for the last few months on our own department’s business continuity planning, availability stats, and general architecture needs. Of course, in that domain, redundancy is some of the core of what IT must provide – redundancy of a form that ensures the ongoing availability of critical services – redundant hardware, network connectivity, power, air conditioning, etc.”

Mr. Kass didn’t argue with the column’s general principle, but did suggest that the continuity thought process is relevant for staffing, and in particular for cross-training and succession planning. To which I say … after all, one of the basic principles of information technology is to plan redundancy. It’s as valid when choosing managers as it is with information engineering.

The main point is that the person in charge has a huge impact on performance. Twice. A week ago it was all about you. If you’ve been in charge for any length of time, and nothing is changing or the situation is deteriorating further, go back and read it.

Here’s the other half: Leaders at all levels can cause or encourage poor performance, which cascades in a sort of chain reaction through the chain of command. This makes it … what? A chain of command reaction? A bad leader recruits bad leaders or prevents good leaders from leading well. Those bad or prevented leaders do the same in turn. And so on, and so on, and so on.

In trying to sort out which of your managers can’t be rehabilitated, an excellent source of information is the collection of performance reviews they’ve filed. When you review them, you’ll almost certainly find some managers who have nothing but problem employees reporting to them. Put these managers at the top of the list of managers who need to find other opportunities — with your competitors if at all possible.

But keep the people who reported to them. I’ve lost track of how often bad managers hide and suppress the abilities of the good people they supervise.

Very recently I heard yet another example — the manager of a small workgroup that consistently hired outside consultants whenever difficult work was assigned, because her direct reports just didn’t have the skills. The recently hired CIO terminated her, asking her workgroup to take on a couple of challenging projects. You know what’s coming: They stepped up to the plate and hit, if not home runs, at least solid singles and doubles — good employees made bad by their manager.

You’ll also find managers who have only stars and superstars on their teams. Some really might be excellent leaders who recognize and insist on talent and drive among those they hire. You can recognize them through the performance reviews they authored. You’ll read nuanced (there’s that word again) accounts of each individual’s strengths and how they contribute to the overall health of the organization. And even the strongest individuals will have received suggestions for areas that would benefit from additional focus and attention.

But most managers with only stars and superstars will have written, over and over again, “Doing fine — keep up the good work.” They might have low standards. They might simply be afraid to lead, erroneously using the approval of those who report to them as a gauge of their abilities. Or, they might have misplaced their loyalty, figuring they’re competing with their peers for the raise pool, and are responsible for obtaining as much of it as possible for their team. (In some circumstances, this loyalty isn’t misplaced, either — it’s required to avoid having their peers raid the raise pool, leaving their employees high and dry. If that’s been the case, don’t blame the managers. Fix the circumstances.)

Keep every strong leader you possibly can. You’ll need them all. Can you turn the others around? They’ll take a lot of your time and attention; many will fail anyway.

Give them the attention. But not a lot of time.

Those who have the potential will demonstrate it quickly given an environment that encourages it. The rest? Give them time and they’ll figure they have all the time in the world.

Which is, of course, just another symptom of the problem.

In boot camp, the military first starves recruits to get rid of their flab. Then it feeds them to build muscle.

Or at least, that’s what I’ve read in various fictional accounts. Not having had the pleasure, I can confirm neither the truth nor the efficacy of this technique. It’s remarkably similar to one I don’t particularly like, but whose efficacy I must admit, for shaking the dust off a complacent organization: Shrinking it through layoffs, then gradually adding the staff necessary to create a newly effective organization.

Among the hallmarks of a complacent organization is the accumulation of second-rate and third-rate employees. You know who I’m talking about. Second-raters just barely pull their weight. Third-raters don’t even do that, but have perfected the ability to hide behind the few strong performers who produce most of the organization’s results. How does this happen?

Complacent leaders hire placeholders who, on paper, have the right skills and who, in their interviews, make it clear they won’t rock the boat. Hiring happens quickly so the hiring manager can go back to sleep. Then the few strong performers, who already have a well-developed bad habit of carrying everyone else, will carry the new faces as well. It’s just business as usual.

And in you come, ready to get your motor running and head out on the highway. What are you going to do? Surgically remove the second and third-raters one at a time?

Which is why the starve-then-feed method is both popular and effective. It’s quick, and it avoids the overwhelming effort needed to document termination-worthy performance one employee at a time. Which in many cases won’t be possible, as many of the employees who need replacing perform at a level that’s just good enough, and report to managers who are incapable of distinguishing good enough from better.

This makes the starve-then-feed gambit hard to argue with. It’s distasteful, but sometimes it’s the only practical tactic.

If you’re in this situation and see starve-then-feed as your only practical option, please … retain and lay off based on merit, not seniority. The whole point is to raise the bar after all. Make it your choice, not an opportunity to volunteer for a generous severance package.

But take the following steps before you do. They’re essential to making sure the organization survives the experience:

  • Develop effective means for communicating directly with the IT staff — listening to them, informing them, and persuading them. Don’t rely on your chain of command. It’s a big part of the problem. It certainly won’t transmit ideas in either direction with high fidelity.
  • Define the desired change in culture, and start the process of culture change. You need a clear picture of where you are now, and a precise description of the change you’re about to undertake. If you don’t, you’ll find yourself overwhelmed with case-by-case decision-making, which inevitably you’ll make in inconsistent ways that cause you to undermine yourself. When you can articulate what you want, start the program. Everything that’s going to happen must fit the design of the desired culture, not the old one.
  • Replace your problem managers. There’s a reason the organization has accumulated so many ineffective employees — the people who hired them and kept them on. Leave the same managers in place and they’ll simply replace the old group of ineffective employees with a whole new group of employees just like their predecessors, except the new ones also lack deep knowledge of your company and its systems.And by the way: One of the most important reasons for developing direct communication channels with staff is that you’ll be replacing many current managers — the existing chain of command.
  • Preserve critical staff. Remember those few strong performers who are carrying everyone else? You can’t afford to lose them. Identify them, develop personal relationships with them; promote those who are looking for that kind of opportunity; give them raises; and let them know — overtly and without any tiptoeing around the subject — that you consider them essential to the organization.
  • Give everyone a chance, but not a lot of breathing room. Managers first: Some have been itching for a chance to lead well but have been stymied by the old leadership and culture. That goes for the staff even more. They deserve a chance.It’s a fine balancing act, though. If you give them too much of a chance, you’ll find yourself reinforcing the old culture instead of promoting the new one.

On the other hand, if you don’t give them any chance you’ll unnecessarily jettison important skills, organizational knowledge, and loyalty. Replacing that will cost a lot. Not only that, but by providing opportunities to succeed, you’ll have fewer sleepless nights yourself.

Think of it as enlightened self-interest.