Stop trying to fit workers in a "good" (or "performer") or "bad" ("underperformer") binary classification and instead adopt a richer language which fits reality of people better.
The problem is illustrated by this HBR essay submitted by the top manager of some company:
When to Fire a Top Performer Who Hurts Your Company Culture
In the essay, the author argues that there's something like "the culture" in a company that needs protection from "zombies" and "vampires" -- the author's words for employees who are damaging "the culture." In contrast to these monsters, there are "stars" and "potentials," which are the good guys.
Needless to say, this binary clasification of "good" and "bad" people is from the standpoint of the company management or owners. But the question is, where do all these "bad monsters" come from?
My (unproven, but quite obvious) thesis is simple: "good" people are, essentially, people who conform, and "bad" people are essentially people who don't conform and can't persuade -- or aren't in the persuasion game -- managament to adjust their goals to accomodate what they think is a better way to do things.
An "unaligned" worker is a worker which could be, previously, a "high performer." Anybody who is really a high performer is a strong candidate to notice something that should be changed about how the business itself works, and if their opinion of that clashes with management, then management will feel threatened.
At that point, the "high performer" which, the author of the above issue claims, is now a "vampire," that is "trying to start a revolt," has to be let go. That is probably true. Lack of cohesion is probably more deadly for a company than mere obsolescence. The marketplace seems very tolerant of buying incremental versions of the same -- to quote Top50 economist and HBR blogger Umair Haque -- "designer diapers".
So there's a basic conundrum which, I think, is not a real "problem" to be solved but simply an interesting facet of the nature of doing business as we currently do: a company is a fundamentally competitive unit, and competitive units need, above all, internal cohesion. Innovation disrupts that. Divergence of ideas, variance, disrupts that. People trying to sabotage the company disrupts that. Burnt out people disrupts that. There's no question about it. However, the other side of this is that if you get obsessed with cohesion, your company will eventually die. Sure, some companies are not interested in not dying: once the business stops making sense, the workers are all fired and the CEO is already rich anyway, and he can create another company from scratch instead of trying to mutate the one he already has into the new one. I'm assuming we're talking about the other type of company -- the one that wants some degree of innovation, and to last -- so let's keep going.
Leaders who think in terms of "good" and "bad" people are thinking in simplistic terms. They are blind to the nuances within these categories. They are simply seeing who to keep and who to fire, and they are thinking unidimensionally in terms of cohesion. All the innovation will have to be personally "sold" to them before it happens; they will centralize all of defining and exploring what the business can be. This can work for some corporations, and some workers. Creative workers who can imagine things but that aren't particular fond of begging for people to understand what they themselves barely believe in, will probably not tolerate such an environment.
Within "performers," there's people performing to metrics and people doing real work. Within the "monsters," there's people who have been corroded from the inside by being systematically played down by the system put in place, while at the same time being given the illusion that their ideas are being considered or that there is a path to getting them to have some effect. And, of course, after some time trying to integrate divergent ideas into the whole, and being faced with the inability to integrate such ideas, a schism will start to develop, and some people will start to "vampirize" others into their ideas. If coupled with the inability to see that it is simply time to go, the result is an attempted "takeover," as the author of the above essay observes. So yes, when it comes down to that, it is better for the holder of the divergent idea to start their own company.
The solution is simple: leaders and workers alike should stop unconsciously buying into the simplistic (and therefore seductive), crass narrative of "performers" and "non-performers," of "good" workers and "bad, bad" workers. Up your game. Up your language. Up your perception and your modeling of reality. Understand that a business is about cohesion. You don't have to villify people to be able to justify to yourself firing them. Be at the same time more respectful and firmer.
As a leader, you have no responsibility to give people "several chances" to "prove themselves." If a manager came to me and said, "I'm going to give you another chance," I would be furious (of course, I am now more able to see when they *think* like that but won't say it because they know I have pride, and if they want to "farm" my capacity a bit more, they have to hide it -- the mask of politeness that manipulators wear). Shove your chance; I don't have to prove myself to anyone. If you don't want my services, I'll take myself somewhere else.
Workers that have self-respect don't want "chances" from "superiors." The meaning-seeking worker wants either on or off. Let them know as early as possible what is it that you want, and if you don't know what you want, admit it to them so that they can stop wasting their times working for a weak leader. Save yourself that trouble, save workers that trouble, and instead be explicit about what you want and what you don't want out of them. Don't project your own insecurities and your own inability to judge whether an idea is good or not into people being "good people" or "bad people." Don't say you're giving people "a chance to prove whether their innovation is worth it" or whether "they're deluded" or "trying to fool you." Instead, know that innovation is exploration, and that there's no such thing as "I'm giving you the privilege of trying out your idea," but instead, "We both don't know where it will lead, but I choose to want it, and therefore, you as my paid worker, will be aligned to the company as you seek it."
Don't lord over people. Don't be arrogant. You're a leader of highly talented people holding PhDs who chose to work for you, not a king of helpless subjects. You have a technical knack for being a player in a marketplace, same as the technical knacks that your "underlings" have. The only difference is that you control the money and they don't. Get over yourself. You can use a language of equality and respect, a neutral, realistic language that doesn't divide people into the "playas" (you, leader) and "replaceable pieces in the chessboard." Start here and you will have made one solid step towards getting this "innovation thing."
Unknown. My belief is that without this basic problem of respect and language, it is too difficult to speak of divergence, variance and innovation. It is too difficult to solve the problem of variance versus the need of cohesion in a company that has brickheaded people as leaders.
First, workers should stop allowing themselves to be emotionally manipulated and manipulated by "leader-talk."
Second, having run out of "top performers" who somehow are now aware of the manipulation game as well as having been trained into PhDs, leaders will now have to admit they are as human as the people they hire, and they will have to adjust their worldview to be able to radiate the kind of fundamental respect that's going to attract and keep these people around. And, paradoxically (not really) they will have to get better at firing people more often as they will be forced to be more honest with themselves about whether they really know what is it that they want to provide the world, of if they are just another manipulative, aimless "leader" who is really purposeless, meaningless to the world and just wants "more money" -- and as such has no technical basis to judge the merit of "alignment" of any innovation but instead will rely solely on whether he knows how to sell it or not.
Dilbert cartoon and that essay where a Pixar manager took all the "zombies" and "vampires" from all these other teams, and then made a team that created one of Pixar's best hits.