Changing peoples behavior is hard. Things that don’t work well: just have more self-control, put stricter punishments in place, nag repeatedly.

Need to attack on three fronts simultaneously:

There are a lot of interesting, in-depth examples in the book. I’m not going to type them all out :)

Example: Donal Berwicki wanted to reduce rate of death due to accidents in medical procedures.

Started with a completely clear goal:

Here is what I think we should do. I think we should save 100,000 lives. And I think we should do that by June 14, 2006 - 18 months from today. Some is not a number; soon is not a time. Here’s the number: 100,000. Here’s the time: June 14, 2006, 9am.

Brought family of patients who had died due to errors to give speeches:

I’m a little speechless, and I’m a little sad, because I know that if this campaign had been in place four or five years ago, that Josie would be fine… But I’m happy, I’m thrilled to be part of this, because I know you can do it, because you have to do it.

Focused on six interventions that were simple to implement and had the potential for a large impact. Made it easy to join the campaign - hospital CEO just has to sign a one-page form to get started.

Estimated number of deaths avoided was 122,300.

Direct the Rider

Find bright spots - places that are already working. Eg fighting malnutrition in Vietnam - look for villages with low rates of malnutrition and see what they are doing differently. Advantage of having an already implemented example to work form, and comes from inside the community so faces less resistance or cultural surprises. Gives clear direction - do what they are doing. Counters the tendency to focus on what’s going wrong, blinding you to potential solutions.

Script the critical moves. Making decisions depletes mental resources, eventually leading to overload and paralysis. Give people clear frameworks, flowcharts, defaults etc to avoid the paralysis. Eg railroad CEO created list of four triage rules to bootstrap out of bankruptcy: only invest money in projects that produce short-term revenue, always pick the solution that is cheapest up-front (even if it costs more later), refer quick short-term fixes to slow long-term fixes, reuse or recycle materials instead of buying new. A simpler example is setting a specific time to carry out a new habit you are trying to create, rather than just trying to do it ‘regularly’.

The book keeps referencing the food pyramid as an example of poor behavioral change, so I’ll point out the Precision Nutrition meal guide as a example of how to do it right.

Point to the destination. Provide an inspirational destination rather than focusing on what’s wrong with the current state. Bad: “increase production rate” or “lower defect rate”. Good: “democratize the automobile”. BHAG - big hairy audacious goal. There seems to be a clear danger of killing motivation if the goal is too hairy. Focusing on the successful case studies is a clear survivorship bias.

Set black and white goals to avoid the possibility to rationalizing your way into failure. Bad: “drink less wine”. Good: “no wine ever”. No wiggle room in the latter. Eg BP announcing “no dry holes” rather than “increase success rate of drilling”.

Motivate the Elephant

Find the feeling. Combine intellectual arguments with emotional appeal. Eg rather than “we lose money because our factories don’t standardize equipment purchases” go for “here are 16 tables piled high with examples of all the different kinds of rubber gloves we buy” - immediate impact.

Tailor message to create an appealing identity / self-image. Eg chemo isn’t the cause of sickness, “it’s the weapon you use to get your life back - take the pills and you can stop being a cancer kid forever”.

Rule of thumb: negative emotions are useful for inducing quick, specific actions whereas positive emotions are useful for driving long-term effort.

Shrink the change. Find the smallest possible starting step, and build up momentum from there. Eg rather than deciding to stop being a slop and live in a perfect palace of cleanliness, decide to spend five minutes after lunch every day cleaning. I set a rule that I have to go outside first thing every morning, which has led to having a regular exercise routine. The trick is that when I wake up I feel shitty but will grudgingly agree to go outside just for a minute to satisfy the rule, but once I get outside in the sun and fresh air I start to feel like moving. Now that I think about it, it’s kind of strange to have a routine that’s built around successfully lying to myself every morning.

Grow your people. Create and reinforce the identity that will lead to the behavior you want. Commitment / consistency effects can be abused to this end. Create the expectation of short-term failure to make motivation less fragile. “This is going to be challenge. There will be setbacks along the way. We will ovecome them.”

Growth mindset blah blah blah.

Shape the Path

Place reminders and nudges in the environment. Eg giving nurses special vests to wear when giving out medication, so that everyone knows not to distract them. I worked at a startup that happened to have blue floor-level lights in the office, so they decided to have ‘blue days’ twice a week. When the lights are on, the main office is silent and the chat room is offline. Helped carve out distraction-free periods for deep concentration.

Remove unwanted temptations and make dangerous behaviors impossible. Eg industrial machines that require two high-up buttons to be pressed before they activate, ensuring that both hands are well out of the way.

Action triggers - specify the precise environment (time, place etc) that will trigger the action you want to remember to perform eg “I will do this homework sitting at my kitchen table after finishing breakfast”. Both a reminder and an artificial deadline.

Create habits that promote the behavior you want eg eating two cups of soup per day as part of a diet, to increase feelings of fullness. Similarly, create rituals to help set up particular mindsets eg morning routine of assembly etc in schools helps switch kids into school mindset.

Use checklists.

Rally the herd. Use peer pressure eg surround the worst offenders with good examples. Place highly visible confederates to kick off behaviors eg find a few people in your company who behave well and make their success visible, so that it feels like everyone else is doing it. Report peer behavior when it isn’t visible eg hotels leaving signs saying “most guests reuse their towels” leads to more towel reuse than signs asking guests to reuse towels. Create fake peers if necessary eg Winstens successful campaign to have as many TV show as possible mention the concept of a designated driver in the dialogue, eg creating a despicable sugar daddy character on Tanzanian radio whose name became an insult.

Keep the switch going. Reinforce desired behaviors with rewards, rather than punishing undesired behaviors. Begin by rewarding an improvement, and then slowly raise the bar.

(There is a final chapter on examples of typical obstacles to change and how to overcome them, which is effectively just more case studies.)


I don’t know what to with all of this. It’s the same reaction I have with all the Heath books, where it all sounds like reasonable, self-evident advice but I have no way to judge it.

Probably the only thing I will attempt to remember is the breakdown into direct the rider + motivate the elephant + shape the path, as well as the idea of scripting critical moves. I tend to forgot one or more of those.