I don't think he meant let those people hang around being disruptive. Just don't structure policy around them. If they are outliers in that respect, then will any policy you come up with make those people productive? Will it have overall positive benefit or will it damage the ability for the other 95% to work?

If you have 95% of your workforce who want to succeed and 5% of your workforce who are "slackers', you can provide tools to help those who want to succeed and fire the slackers. On the other hand, you can force through workflow policies that force the slackers to do minimal work while frustrating and slowing down the workflow of the rest of the staff who want to excel.

As well, the more freedom you give to those staff who want to excel, the more rope you give those staff who will not to hang themselves with. Force them to jump through hoops to show that they are working and they will do what they can to hide the fact that they're not. Give them free reign to slack off and they will not have false metrics to hide behind, nor will they feel the need to hide so much.

I think that he means do not do things like create a process that everyone must follow, full of red tape, simply because your worst employees cannot perform a task correctly.

And you should still not focus on them! That is a statistical fallacy.

Improve the system, and everything improves. Don't focus on outliers—they will always exist. You'll be killing individual ants when you really want to go after the queen.

Sure, you can always deal with individual cases as they come up. You can't ignore that there are differences between the performance and fit of different employees. So hire and fire the best you can.

The key is what happens after that. You'll still have a rough bell curve of how people perform. You can keep chopping off the bottom of that bell curve, but the problem is that that practice is in itself demotivational. It's bad for culture, it's bad for your company. Basically it's negative reinforcement for all your employees, and the response to negative reinforcement is—to put it lightly—not optimal.

Your bell curve will be sucked backward by the firing black hole and people will start acting out of fear and individual motivations, which include: 1) not being fired, 2) keeping their job at all costs, and 3) looking better than everyone else. Rarely does the fear response motivate people to actually work harder. Instead it leads to a culture of in-fighting, blame, and ladder-climbing games. And in no part of this story are the overall goals of your company recognized, let alone respected or achieved.

This is why you don't concern yourself with the slackers. If they really are slacking and not getting anything done, you may fire them—no one is stopping you. But this is not your focus. Your focus should be on moving the entire bell curve up by 20%, motivating people to perform beyond their basic abilities, and empowering them to work. These are fluffy phrases, but they are also true. You start from the top down, make sure everyone understands the overall goals of the company, make sure they are well educated about the means and purpose of their work, and that they are enabled to work as well as they can. This is the essence of W. Edwards Deming's philosophy, and yes, it does work this way.

Because, more often than not, your "slacker" is not simply a bad person who can't work. Very few of those exist in reality. He is probably a person who has not learned the right things, encountered the right motivations, or understood the true purpose of his work, or he has confounding factors preventing him from reaching his true abilities (see Maslow's basic needs theory). Most people want their work to be meaningful, because people fundamentally want their lives to be meaningful.

This is what I'm saying: it's far more effective to improve the system surrounding your employees than to simply treat your employees as unaffected robots whose performance is static. Again, I'm not saying you can't fire people when they don't work: I'm just saying that it would be dumb to only focus on that strategy as though there were no other variables, and in fact, the system variables are the important ones when you actually look at it.

Best business advice of the last few hundred years:

"If you want to build a ship, don't drum up people together to collect wood and don't assign them tasks and work, but rather teach them to long for the endless immensity of the sea."

-- Antoine de Saint-Exupery

What does the size of the company matter? If 5% of the workforce in a 100 person company slacks off, vs 5% in a 50k person workforce, why is should policy be any different for the 50k workforce?

If you're implying the percentage of slackers scales non linearly as the size of a company grows then I'd love to see some research to back that up, as I'd be genuinely interested if that were the case.

One possibility is that as the number of slackers grows, the amount of damage that they can do increases at a greater-than-arithmetic rate, particularly if they tend to cluster together.

For example, if 5% of a company overall is composed of slackers, but 25% of their customer support department are slackers, they may do outsized damage.

So get rid of them. The trick is to get rid of them without enacting policy that negatively affects the morale of your productive workforce.

" Fundamentally, everyone wants to do a good job (the statistical outliers who do not follow this are not worth focusing policy on)"

You have just said you should focus a policy on getting rid of them. I agree with this, but it goes against what the article just said to do. I posit this is because he does not have a large enough company to have enough slackers that he needs to deal with them systematically, and instead can simply deal with them on a one-off basis.

Corporate policies can miss their targets. If a policy targeting the slackers existed, than part of it's job would be to target said slackers, which it would only be able to successfully do with some probability. However, if there isn't a statistically significant number of slackers, the policy could end up falsely targeting more hardworking employees than slackers. It's a classic probability/confusion matrix problem and goes to show that you have to be damn sure that your policy is only targeting the right people.

A company willing to risk harming its honest employees for minimal benefit is not one I would like to work for.

You seem to assume there isn't some minimum non-relative performance bar you can set for what "slacking" is, without harming hardworking employees.

Also, nobody said the policy had to be completely rigid, but having no policy will become a disaster as the company grows larger, and HR actually becomes significant.

Fair enough. But my experience tells me that large corporations + many policies = many missed targets.