Perversely, the desire to avoid risk often magnifies it. Dumping money into big me-too projects with modest upside is a lot more perilous than seeding lots of early-stage ideas that are further out on the fringe. In the age of upheaval, incrementalism is the riskiest bet of all. What’s needed is a radical shift in how we think about experimentation. The goal isn’t simply to reduce the uncertainty around new products or get them to market faster, but to build an organization where everyone is working to extend the boundaries of what’s possible. That’s how an organization buys insurance against irrelevance.

Important questions at this stage will include:

1. What’s our proposed solution, in a single sentence?

1. What are the key components of our hack?

1. What hypotheses do we need to test?

1. Who will participate in the experiment?

1. What data will we collect?

1. How do we ensure we get meaningful results?

1. How much time will we need to run the experiment and what resources will be required?

CONTRARIAN THINKING. If a problem’s been around for a while, it probably can’t be cracked with conventional thinking. Seek out the positive deviants, like Nucor and Haier. Borrow ideas from other domains, like biology, startups, and crowdsourcing. Rigorously challenge your deepest assumptions. Do all this, and you’ll increase the odds of finding a novel solution.

COMPASSION. People aren’t merely skeptical; they’re cynical—and with good reason. Everyone’s fighting their own corner and looking out for their own interests. When asked to help, most people will ask, “What’s in it for me?” To jump this hurdle, you have to put others first. When colleagues see you working to understand their needs, when you help them craft their experiments, and ensure they get the credit, they’ll start to trust you. When your compassion shines through, people will take risks with you and pick you up when you fall.

CONNECTIONS. Building a community is the most important thing an activist can do. This is the ultimate multiplier of individual effort. Employees eager to try something new often make the mistake of asking their boss for permission. Usually they get shot down, or win only grudging support. This isn’t entirely the manager’s fault. A priori, it’s hard to know whether an underdeveloped idea is brilliant or batty. Since great ideas are rare, the default setting for most managers is to say no. So don’t go up, go out. Talk to your peers. Find a few colleagues who will help you build and run an experiment. It’s easy for a manager to say no to a lone supplicant, but much harder to turn aside a small band of partisans who are passionate about making things better and have already made a start.

1. Conduct Tests

The test design process leads to the actual testing phase and the analysis of results. Here,

we recommend taking what we sometimes call the lazy person’s approach to strategy. Simply put, first test the things you’re most dubious about. Take the condition the team feels is the least likely to hold up, and test it first. If the team’s suspicion is right, that possibility will be eliminated without the need to test any of the other conditions. The possibility has already failed an essential test, so no more tests are necessary. If, on the other hand, the possibility passes the first test, move on to the condition with the next- lowest confidence level, and so on...

To generate choice and commitment, companies actually need analysis that is an inch wide and a mile deep—focused precisely on the concerns that prevent the team from choosing and going deep enough in that particular area to meet the team’s standard of proof. That is what reverse engineering enables you to do: probe precisely and deeply into the barriers to choice.

“REVERSE-ENGINEERING DOS AND DON’TS

• Don’t spend a lot of time up front analyzing everything you can; instead, use reverse

engineering to pinpoint only what you really need to know.

• Do frame a clear and important choice up front; make it real and significant.

• Do explore a wide range of where-to-play and how-to-win possibilities, rather than narrowing the list early on to those that feel realistic; unexpected possibilities often have interesting and helpful elements that can otherwise be dismissed out of hand. Learn from them.

• Do stay focused on the most important question (what would have to be true for this to be a winning possibility?), listing the conditions under which this possibility would be a really good one.

• Don’t forget to go back and eliminate any nice-to-have conditions; every condition should be truly binding—if it weren’t true, you wouldn’t pursue the possibility.

• Do encourage skeptics to express concerns at the specify-barriers stage; have them articulate the precise nature of their concerns about specific conditions.

• Don’t have proponents of a given possibility set and perform the tests; ask the

skeptics to do it. If the skeptics are satisfied in the end, everyone else will be too.

• Do test the biggest barrier first. Start with the condition the group feels is least likely to be true. If it isn’t true, the conditions required do not hold and you can stop testing.

• Do use a facilitator to run the reverse-engineering process; it helps to have someone to attend to process and group dynamics as you work through the thinking tasks.