How to Influence Yourself to Action When It's Hard

The full conversation with Rich Diviney covers the swim-test story that changed how he thinks about picking elite performers (yes, a candidate who couldn't swim applied to be a SEAL), the "moving horizons" technique he used to get through Hell Week, why he says we never actually reach our potential, and how he and Andrew Huberman breaks down fear.

Listen to the full episode →

A few years ago, Katy Milkman and Angela Duckworth ran one of the biggest behavior-change experiments ever done. Thirty scientists from 15 universities designed 54 different programs to get people to the gym. More than 61,000 gym members took part.

Fewer than half of the programs (45%) moved the needle at all. Only 8% were still working after the program ended. And independent judges couldn't reliably predict which ones would win.

Read that again. The people who study this for a living couldn't guess what would work.

So if your gut says "I just need more discipline," your gut is probably guessing too.

Here's what the research does tell us. "Hard" comes in three kinds. Starting, keeping going, and keeping it up are three separate problems. Each one breaks for different reasons, and each one responds to different tools.

Let's take them one at a time.

For each “hard”, I’ll list them by Quick Wins, Simple Experiments, Good Investments and lastly Add-On Tools (see below.)

A quick note on how to read this. Treat it as a research-informed guide. Some strategies have strong evidence for a modest effect. Some have exciting ideas behind them and very little direct testing. And the best strategy is always the one that matches your kind of hard. A quick win for starting won't help much if your real problem is maintaining something hard.

This is the initiation problem. You know exactly what to do. Write the article. Make the call. Go for the run. And somehow, you do something else.

A few things are usually going on. The reward is far away and the discomfort is right now. The first step may be fuzzy and something easier is within reach.

Quick Wins

If–then planning. Peter Gollwitzer named these implementation intentions. Instead of "I'll work on the proposal this week," you decide when, where, and how: "If it's 7:30 on Monday, then I'll open the proposal and write for 10 minutes."

This is one of the best-studied tools in all of behavior change. Gollwitzer and Paschal Sheeran's 2006 meta-analysis of 94 tests found a medium-to-large effect on reaching goals. In 2025, Sheeran and colleagues analyzed 642 tests and found effects ranging from small to medium-large depending on what was measured. The plans worked best when they used a true if–then format, when people actually cared about the goal, and when they rehearsed the plan.

Why does it work? You hand the decision to the situation. When 7:30 Monday arrives, your brain recognizes the cue and the response is already loaded. There's nothing left to negotiate.

Application:

WOOP. Gabriele Oettingen spent years studying positive fantasies, and found that daydreaming about success can actually drain the energy you need to chase it. Her fix is WOOP: Wish, Outcome, Obstacle, Plan. You picture the good outcome, then name the inner obstacle most likely to stop you ("I'll feel tired and scroll instead"), then make an if–then plan for it.

A 2021 meta-analysis of 24 effects and nearly 16,000 people found a small-to-medium effect on goal attainment, likely a bit smaller after adjusting for publication bias. Not magic. But it takes five minutes, and it's honest about the part most plans ignore: you.

Simple Experiments

Make the first step tiny. One sentence. Shoes on. Open the deck. This one makes intuitive sense, and it borrows from the research on small, near-term goals. I'll be straight with you, though: I couldn't find strong direct tests of the "just do the tiny version" technique on its own. Treat it as a cheap experiment.

Pick a fresh start. Ever notice how packed the gym is in January? Hengchen Dai, Katy Milkman, and Jason Riis named this the fresh start effect. At one university gym, the odds of a visit jumped about 33% at the start of a week and 47% at the start of a semester, and rose after people's birthdays. On stickK, a site where people sign goal contracts, new contracts more than doubled at the start of a new year. New beginnings seem to put the old, imperfect you in the past.

Most of that evidence comes from watching what people already do, so it shows a pattern. It can't prove that picking a date causes the change. But one experiment points the same way. When more than 6,000 university employees were invited to start saving for retirement on a "fresh start" date, like New Year's Day or their birthday, about 50% more of them enrolled than when the date was simply a few months away.

You've probably heard that "I'll start Monday" rarely works. Both can be true. Monday as an excuse to skip today is procrastination in a nice outfit. Monday with a plan attached is a real starting line. If you can start today, start today. If you're going to wait anyway, make the date mean something.

Application:

Good Investments

Change the room. Clothes beside the bed. The doc already open. The phone in another room. Environment design is a huge category, and the evidence varies a lot by type. Defaults and making the good choice the easy choice tend to hold up best. Pure "here's some information" nudges hold up worst. One honest note: a big 2022 nudge meta-analysis was later challenged for publication bias, so be skeptical of anyone who tells you nudges always work.

This is the persistence problem. You started. Now it's uncomfortable, progress is slow, and you can't see the end.

Quick Wins

Moving Horizons. Rich explained it to me this way. In any situation, your brain is trying to figure out three things: duration (how long will this last?), pathway (what's my way through?), and outcome (what's the end state?). When you can't answer one or more of those, stress goes up.

So you create your own answers. Rich's version on the beach: "Duration, from now until end of sand berm. Pathway, from here to end of sand berm. And outcome, end of sand berm." Get there. Notice it. Pick the next one.

He also gave me a line I wrote down: "The past is history, the future is fiction." Anxiety lives in the fiction. The horizon keeps you in the part you can actually control.

I've done this without knowing it had a name. I grew up in Singapore, where military service is mandatory. In one exercise, they blindfolded us and dropped us into the deep end of a pool. I remember thinking, very clearly, I'm going to die here. Then, almost by accident, I focused on one thing. Can I reach the bottom and push myself up? That was it. That was my horizon. (I still don't love deep water, by the way.)

Here's the honest science part. Moving Horizons is Rich's name for a near-term goal you set in the moment, and the near-term goal part is well researched. In Bandura and Schunk's 1981 study, kids with small session-by-session goals finished 74% of the math material, compared with 55% for kids with one big distant goal. They also became more persistent on hard problems. The distant goal, in the authors' words, "had no demonstrable effects." Rich's in-the-moment version, with its duration, pathway, and outcome, hasn't been tested directly. But it stands on solid ground.

Application:

Simple Experiments

Temptation bundling. Katy Milkman's idea: only let yourself have a "want" while you do a "should." In her first study, gym members who could only listen to gripping audiobooks at the gym went about 51% more often at first. The effect faded over time, especially after Thanksgiving. A much larger follow-up with almost 6,800 people found a smaller lift: about 10–12% more workouts per week. Modest, but real.

Good Investments

Make progress visible. Benjamin Harkin and colleagues analyzed 138 studies with almost 20,000 people. Getting people to monitor their progress helped them reach their goals. Two details matter. The more often people checked, the better. And the effect was even stronger when progress was physically written down or reported to someone else.

One more nuance worth knowing: tracking the behavior changed the behavior, and tracking the result changed the result. If you want to write more, count sessions. Counting followers measures something else entirely.

Add-on Tools

Put something on the line. In a study in the Philippines, Dean Karlan and colleagues offered smokers a savings account. They deposited money, and if they failed a nicotine test after six months, they lost it. Only about 11% signed up. Those offered the account were 3 to 6 percentage points more likely to pass the test. Most still didn't quit. Commitment devices can work, but they're a tool for specific moments, and most people don't want them.

This is the consistency problem. You've done it. You know you can. Then the novelty fades, the calendar fills up, and it slips.

This is the hardest of the three. Remember the gym megastudy: plenty of programs worked for four weeks, and almost none kept working after.

Simple Experiments

Habit stacking. "After I pour my coffee, I write one paragraph." Habits form when the same cue keeps triggering the same behavior. That's why the cue matters so much.

In a small study on flossing, Gaurav Judah and colleagues had people floss either before or after brushing their teeth. The after-brushing group built a stronger habit, and eight months later they were flossing more often. It was a small study, so treat it as a promising signal. But it captures the idea: hook the new thing onto the end of something you already do.

Plan the comeback. In the gym megastudy, the single best program gave people a small reward for returning after they missed a workout. And in Phillippa Lally's habit study, missing one day didn't derail anyone. One missed day is normal. What matters is how fast you come back.

Application:

Good Investments

Let your reps build the identity. "I'm someone who writes." "I'm someone who keeps promises to myself." This is popular advice, and it rhymes with real psychology. Daryl Bem's self-perception theory suggests we often figure out who we are by watching what we do. Daphna Oyserman's work on identity-based motivation shows structured programs that connect identity to action can change behavior.

But here's the catch. I couldn't find good evidence that repeating an identity statement on its own changes much. One famous study found that asking people to "be a voter" boosted turnout more than asking them "to vote." A larger replication didn't find the effect. So the order matters. Do the thing, then let it count as evidence of who you are.

Add-on Tools

Borrow an identity for the hard moment. There's a fun twist on this. In 2017, Rachel White, Angela Duckworth, Ethan Kross and their colleagues gave 180 kids, ages 4 to 6, a boring 10-minute computer task. An iPad sat nearby, tempting them to quit. Some kids were asked to think about themselves: "Am I working hard?" Others got a prop and became a hard worker they admired, like Batman or Dora the Explorer: "Is Batman working hard?"

The Batman kids stuck with it the longest. Six-year-olds playing a character stayed on task about 55% of the time. Six-year-olds thinking about themselves managed about 35%.

Researchers call it the Batman effect. Performance coach Todd Herman wrote a whole book on the grown-up version, The Alter Ego Effect: stepping into a character so the hard thing feels less like a referendum on you.

Two caveats. That study was with young children on one task, and no one is sure exactly why it works. The closest adult research comes from Ethan Kross on self-distancing: people who coached themselves by name ("Howie, you've got this") instead of "I" felt less anxious before giving a speech. So treat an alter ego as a tool for getting through the hard moment. The reps are what make the identity yours.

Application:

References HERE