Recently, I wrote a list of “12 Things Good Bosses Believe.” Now I want to delve into the ninth belief mentioned in that article: “Innovation is crucial to every team and organization. So my job is to encourage my people to generate and test all kinds of new ideas. But it is also my job to help them kill off all the bad ideas we generate, and most of the good ideas, too.”

An evidence-based mantra is that, to get a few good ideas, you and your colleagues need to generate a lot of bad ideas. I wrote about this notion in my article “Forgive and Remember” where we saw that, to yield a dozen or so commercially successful ideas for toys, a group at IDEO generated more than 4,000 nonstarters. It turns out, however, that the best managed enterprises don’t just recognize the flowers among the weeds; they mow down a lot of the flowers, too.

I first started thinking about this about five years ago, after a conversation with a Yahoo! executive participating in the Customer-Focused Innovation program that Huggy Raoo and I run at Stanford University. Yahoo! had just invited Steve Jobs to address their top 100 or so bosses. Jobs advised them that killing bad ideas isn’t that hard—lots of companies, even bad companies, are good at that. He insisted that what is really hard—and a hallmark of great companies—is killing good ideas. For any single good idea to succeed, it needs a lot of resources, time, and attention, so only a few ideas can be fully developed. The challenge is to be tough enough to do the pruning so that the survivors have a chance of being implemented properly and reaching their full potential.

Since then it’s also become clear to me that good product and experience design depends on tossing out most good ideas. If too many of them are thrown in, then the result is a terrible and confusing Frankenstein of an offering. (This seems to be many people’s objection to Microsoft Word: It does everything, so therefore is annoying and confusing to use for many single things.)

The implication, then, is that the “innovation funnel”—where a lot of ideas are whittled down to a precious few—should contain two major filtering stages: One is when you get rid of the bad ideas; and the other is when you throw out good ideas that aren’t quite good enough to justify a thinner spread of resources, a greater diffusion of focus, and possibly a more complex customer experience.

If you take this argument to its logical conclusion, it means that a great boss—and let’s define that as a boss whose team delivers innovation—might track the following two metrics:

1. How many good ideas are killed? If this number isn’t high enough, that is a bad sign. It means either that not enough ideas are being generated, or that important hard choices aren’t being made.

2. How many people are complaining—even leaving—because of good ideas being killed? This really is what makes the pruning so hard. It’s tough on the people who come up with ideas and are emotionally invested in them. Being the direct cause of their complaining, and even departure, is awful—and certainly doesn’t make you feel like a great boss. But if no one is complaining, that’s a worse sign. This kind of frustration is an unfortunate byproduct of an effective innovation process, and if your people don’t have enough pride and confidence to get upset when their innovative ideas are killed, then something is wrong with them—or your culture.

These are weird metrics, but they make sense given Jobs’ argument. His argument also resonates with our experience teaching in the Institute of Design at Stanford and my experience working with creative teams in industry: The groups that often do the worst work have too many pet ideas and can’t bring themselves to kill enough of them, so they don’t do a decent job on any of them.

Groups that can’t kill enough ideas often suffer from bad group dynamics, either because multiple members won’t allow the group to kill their pet ideas, or because the group avoids difficult conversations and decisions. When we advise the creative teams we coach at Stanford and elsewhere, we always stress that this “Sophie’s Choice” point in the process will come, and will have to be well-managed. The team will have to be prepared to kill ideas it has nurtured and come to love.

To put this in a broader context, although the words “creativity” and “fun” are often used together, there is a lot about doing creative work that is no fun—failure, confusion, and conflict are par for the course; and even when you are doing things right, it involves killing good ideas and making people angry. That’s why, as I compiled my set of Good Boss, Bad Boss distinctions, this one made the cut. Enterprises depend on bosses to manage innovation as well as implementation. The very best bosses teach and inspire their people to accept defeat gracefully and move forward to implement the selected ideas, even if none of their pet ideas made the cut.

This article was originally posted on The Conversation blog of Harvard Business Review .