Evil Genius 101

Tom Limoncelli has taught many excellent courses at LISA over the years, so when I saw he was teaching a new course called "Evil Genius 101" I was intrigued. Fellow blogger Matt Simmons described the course as "how to win friends and coerce people", and that was fairly apt. Okay, "coerce" is probably a little strong.

Limoncelli defines evil as "the desire to make big changes to the status quo" and genius as "someone who works smartly". The general idea of the class is that you're a person who wants to make big improvements to your workplace and you want to get management, coworkers, and users to go along with your plans. The first step is to develop your evil plan by using the evil acronym EASI: Enumerate, Assess, Select and Implement.

In the enumeration step, you develop a list of your high-level responsibilities and the associated taste. Even if you stop here, that's already an important accomplishment, but you can be more evil. To assess yourself, you have to figure out how good you are at each of the tasks. Limoncelli drew from the Capability Maturity Model and Gene Kim's "the three ways of DevOps" to identify five milestones and the associated characteristics. Now it's time to select something to improve; it doesn't have to be the lowest-scoring task, but should be the one with the biggest impact. Devote energy fixing the bottleneck, because everything else is an unhelpful local optimization. Finally, you implement your improvement plan. This is largely dependent on convincing people that your plan is good.

The second half of the course focused on people. They don't like change, for a variety of reasons. Whatever reason a person has for not liking a change, they will have an emotional response to it. Winning support for your plan requires anticipating the response and crafting your pitch to address it. Some people respond well to logic and data, others require the endorsement of a trusted party or an authority. Still others will be more likely to go along if you can appeal to emotion.

There are two main strategies for making large changes. The first is "splinter, prove, and grow": splinter off a small group to make the change, prove that it works, then keep growing the splinter group until it reaches 100%. The second is related, except instead of splintering the group, you're splintering the plan. "Hide the full plan" involves only revealing one step of the plan at a time. This helps to reduce the target for people to fight against, as well as simplifying what you have to explain.

Limoncelli then addressed strategies for convincing executives, managers, coworkers, and users. Executives need simple, non-technical messages that focus on the business impacts (especially revenue!). Managers also need a focus on business, including statements of undeniable value. Coworkers will go along with a change when it fixes a problem that causes them pain and they trust the person proposing the change. It often helps to convince them that the change was their idea. Convincing users requires developing their trust first. Finally, Limoncelli gave advice on using data to drive decisions and how to present data in a clear manner.

To end the session, each member of the audience had to say one thing they learned in the course that they planned on doing right away. And each attendee received an "Official Evil Genius" certificate, complete with evil rainbows and the most evil font: Comic Sans.