Summary of "How to Change"

Core Idea

Katy Milkman’s central claim is that behavior change works best when you match the strategy to the obstacle, rather than relying on willpower or one-size-fits-all advice.
Her book is built around the idea that people fail for different reasons—temptation, forgetfulness, laziness, self-doubt, and bad timing—and each barrier calls for a different tool.
Change is often easier at fresh starts, when people feel separated from past failures and more open to a new identity.

Fresh Starts, Motivation, and Instant Gratification

The fresh start effect shows up around New Year’s, birthdays, Mondays, semesters, seasons, and major life transitions like moves, promotions, or parenthood.
Fresh starts help because they create a sense of a do-over, making old failures feel less relevant and future success feel more possible.
But fresh starts can also disrupt people who are already doing well, because resets can break momentum and weaken good routines.
Milkman uses sports and workplace evidence to show that these timing effects are real: people are more likely to act on goals when they feel a symbolic reset.
A major obstacle to change is present bias, the tendency to choose immediate comfort over larger future benefits.
One solution is to make the desired behavior more immediately rewarding, as in Stockholm’s piano stairs, where stairs became more popular when they were made fun.
Milkman’s temptation bundling pairs something people want with something they should do, such as listening to audiobooks only at the gym.
The key is that the indulgence must be genuinely restricted to the target activity; bundling works best when the pairing feels real and exclusive.
Gamification uses points, stars, and symbolic rewards to add immediacy, as in the Wikipedia editing experiment where symbolic recognition improved retention.
But gamification can backfire when people do not buy into the “magic circle,” especially if it feels imposed rather than chosen.

Commitment, Planning, and Reminders

When procrastination is the problem, Milkman often recommends commitment devices that block future temptation instead of asking for more self-control later.
Examples include locked savings accounts, self-imposed deadlines, smaller plates, app-based limits, and self-exclusion lists.
Cash commitment devices are especially strong because they put money at risk if the goal is missed, though they can feel strange and many people refuse them.
Soft commitments create a social or psychological cost instead of a financial one; doctors, for example, reduced unnecessary antibiotic prescribing after signing and posting pledges.
Milkman distinguishes sophisticates who know they need constraints from naïfs who overestimate their own willpower, which helps explain low uptake of commitment tools.
Another common failure is forgetfulness, which is often easier to solve than laziness.
Reminders work best when they arrive at the exact moment action is possible, not hours too early.
Implementation intentions—plans in the form “When X happens, I’ll do Y”—turn vague goals into cue-based action.
In studies, cue-based planning improved outcomes for Christmas goals, voting turnout, flu shots, and colonoscopy completion because it made behavior more concrete and easier to remember.
Milkman warns that too many plans at once can become demoralizing, and that for complex goals checklists can outperform memory alone.
Distinctive cues matter: unusual prompts are more memorable than generic ones.

Defaults, Habits, and Long-Term Maintenance

Milkman’s Chapter 5 argument is that laziness is often just a preference for the path of least resistance, which can be harmful or useful depending on how systems are designed.
Defaults harness that tendency by making the beneficial option happen automatically unless people opt out.
At Penn Medicine, switching prescriptions to generic by default raised generic prescribing from 75% to 98%; similar defaults also reduced opioid prescribing and increased cardiac rehab referrals.
Defaults are powerful when the decision can be made once, but they do not solve problems that require repeated action, like exercise or healthy eating.
For repeated behavior, the brain’s default settings are habits: repeated actions in stable contexts become automatic.
Good habits form through repetition, consistent cues, and rewards, and they are often what disciplined people rely on instead of raw self-control.
Milkman stresses that too much rigidity is the enemy: a habit that only works in one exact setting is fragile.
More flexible habits are stickier because they survive schedule changes and real-world disruption.
New habits can be built by piggybacking them onto existing ones, such as flossing right after brushing teeth.
Tracking streaks, like Ben Franklin’s virtue charts or Jerry Seinfeld’s “Don’t break the streak,” helps because it reduces forgetting and makes lapses visible.
Habits are strongly tied to context: the movie-theater popcorn study showed that cues can trigger behavior even when judgment would suggest otherwise.
Milkman’s broader point is that change is often like managing a chronic disease: it requires ongoing maintenance, not a one-time fix.
In large studies, many interventions boosted gym visits while active, but only a portion of the gains persisted after support ended.
The same pattern appeared in home-energy conservation, where gains faded when feedback stopped.
Karen Herrera’s story captures the book’s end state: lasting change depends on a customized system of accountability, planning, tracking, and temptation management that can be adjusted as obstacles change.

What To Take Away

The book’s main lesson is diagnostic: first identify why change is failing, then choose the right mechanism rather than trying harder.
If the problem is timing, use fresh starts; if it is temptation, use bundling or commitment devices; if it is forgetfulness, use implementation intentions and reminders.
If the goal is repeated behavior, build habits and design for flexibility, not brittle perfection.
Lasting change usually comes from systems that anticipate human weakness, not from a single burst of motivation.