Posts Tagged ‘ideas’

Randy Nelson of Pixar: Collaboration is like improv theater

Tuesday, August 10th, 2010

I learned of this video of Randy Nelson, dean of Pixar University, from Garr Reynolds’ Presentation Zen blog. Garr points out many interesting facets of Randy’s presentation, but I was struck by the initial message in the talk, and I’ve been thinking about it a lot the past few days.

Randy says this (all in the first 1:30 of the video):

Two core principles of improv have always guided us. The first is: Accept every offer. If an improviser says to you, “It’s raining a lot in here today,” you don’t say, “Raining in here?” You say, “That’s why they gave us umbrellas.” It’s an offer. You don’t know where it’s going to go, but the guarantee you have: if you don’t accept that offer, it’s going nowhere. You’ve got a sure thing on the one hand: dead end. Or you have a possibility on the other.

The other principle is: make your partner look good. What a great thing. So you know on a team that anything anybody says to you, you’re going to get a chance to “plus” that, you’re going to get a chance to have that be on the table. And they’re going to try to make you look good, not make you look bad.

At Pixar, what we mean by “plussing” is this. You take a piece of work, you take something you’re working on collaboratively, and when it’s given to you, you don’t judge it. You don’t go, “Oooh, this is pretty good; this is what I’m going to do to make it better.” Or, “This isn’t so good, this is what I’m going to do to fix it.” You say: “Here is where I’m starting. What can I do with this? …How do I accept the offer and make my partner look good?”

This made me think of how I learned to collaborate, starting in engineering school and continuing as a programmer and software designer. Collaboration in that environment meant, largely, fighting over ideas. This is how we should design a data structure. We should design a home screen this way. It was a combination of thinking and persuasion. Whoever had a good idea and could persuade the others in the group of its merits, would win the argument.

In my mind, these weren’t destructive arguments. They were, in fact, thrilling. You won some, you lost some, and the results were often really cool.

But when I moved into management this method didn’t work so well. It doesn’t help quiet people contribute to ideas. It’s a peer method and doesn’t work well in a hierarchy. It can divide people. It can be intimidating.

So I’ve been thinking about how to apply the improv model to more situations. When someone comes to me with an idea, how can I, rather than dismissing it initially (which I am prone to do), “accept the offer”? And how can I use my capabilities, not to critique the idea, but instead to make my partner look good?

When innovating, seek out more, and more varied, ideas

Wednesday, November 18th, 2009

innovation tournamentsI’ve been reading the book “Innovation Tournaments” by Christian Terwiesch and Karl Ulrich of the Wharton School. The book sets out a methodology (the “tournament” of the title) for companies to generate and systematically winnow down innovation ideas to eliminate all but the most exceptional opportunities.

Two brief observations:

One, the authors suggest that almost any company’s innovation performance would be helped by increasing the number of ideas going into the top of the funnel. Early-stage evaluation (a la “Discovery-Driven Growth“) is cheap and fast, so the cost of, say, doubling the number of ideas reviewed isn’t significant when compared with an overall innovation budget.

[It was interesting to read today's post by tech venture capitalist Fred Wilson, in which he outlined his approach to finding new opportunities: (1) making public his strategies, ideas, and passions so that entrepreneurs know in advance what he's looking for, and (2) meeting with as many people as he can, every day. In short, a strategy to add lots of opportunities to the top of his funnel.]

Two, along with the sheer number of ideas, the variability of the ideas is important. High variability increases the possibility that a truly outstanding idea is found (given that truly outstanding ideas, like 7-footers with great athletic ability, are few and far between). In that event, increasing the number of ideas coming into the funnel increases the likelihood that a truly outstanding idea is looked at.

Ironically, methodologies like Six Sigma seek to limit the variability of processes. When (mis)applied to disciplines like innovation, they are very successful at impeding the success of the effort.

Related post:
On Discovery-Driven Growth
Processes as art and science