Strategy memo - behavioral design - after Katy Milkman, How to Change

Ladder already
runs on behavior
science. Let's
climb higher.

Katy Milkman's How to Change argues that most behavior change fails because people apply generic willpower advice to specific obstacles. The book maps out a set of distinct levers: social pressure, temptation bundling, fresh starts, commitment devices, and more. Each one beats a different psychological barrier. Milkman's core idea: don't fight low willpower with more willpower. Diagnose the specific obstacle, then match a lever to it. Some of Ladder's biggest behavioral levers are already in place and ready to be pushed further. A few more are half-built and ripe for a polish pass. And some are wide open, unclaimed opportunities.

The frame One obstacle → one lever → one place it lives in the member lifecycle. Grounded in Milkman (2021) + ongoing product conversations.

The diagnostic premise. Generic "stay motivated" nudges fail because they ignore why a given person stalls. Each lever below beats a different obstacle - so they're grouped by where they do their work in the lifecycle. That's also how to prioritize them against whichever metric is weakest.

Where each lever works
A
Activation
Get them in the door and through week one
B
Retention
Keep them climbing past the month-two wobble
C
Resurrection
Bring the lapsed back without shame
A · Activation

Get started

01

Make it easy / defaults

In place - protect it

Beats decision fatigue & friction

Behavioral science is clear: the harder a behavior is to start, the less likely it happens. Remove friction and adherence goes up, almost mechanically. Ladder's coach-assigned plan is a textbook application of this: one decision made for you, one tap to start. Greg's core philosophy for Ladder is "don't think" -- and that's the right instinct. It's the moat. Vibe search, Flex, and build-your-own already live as secondary options, which is the right call: giving members access to more choice without making choice the default path. The ongoing test for any new surface is the same question: does this make a member think more, or less? Push the week into the calendar, pre-download for offline, auto-load today's workout. Every low-friction touch point compounds.

Effort low/Impact high/Defend, don't dilute

02

Cue-based planning

In place - extend

Beats forgetting

The research calls these "implementation intentions": specific if/then plans that dramatically improve follow-through. Ladder's current scheduling is basic: you can set a push notification for a single workout at a specific time, but there's no Google Calendar integration, no bulk scheduling, no recurring schedule for all your weekly workouts, and no way to see that a workout is scheduled after you've set it. Just a push notification. The gap is wide. A proper implementation intentions layer would let members anchor workouts to real-life cues ("after I drop the kids at school," "when I get home from work"), set a recurring weekly schedule across all their priority workouts, see scheduled workouts on a calendar view, and sync to Google Calendar. Cued intentions survive the moment; vague ones don't.

Effort low/Impact med-high

B · Retention

Keep climbing

03

Social accountability & norms

In place - extend

Beats no social pull

"Ladder is better with friends" is the framing that gets to the heart of it. Right now Ladder is a solo game. Squads would make it multiplayer: small, tight-knit groups built around people you actually know, where cheers, a leaderboard, and in-app chat all apply to a tighter circle than the current team construct. The retention logic is compounding (friends who train together stay together) and the referral logic is built in ("we should get so-and-so in here, they're not on Ladder yet"). Phase 1 is connecting with people you already know. Phase 2 is tribes of like-minded people who train the same way, public or private, WHOOP-style. The unused layer on top: descriptive norms. Surface "18 of your 24 teammates trained this week" and you make the behavior feel expected, not exceptional. Leaderboard design matters here: calibrate it to lift the middle of the pack, not demoralize the back.

Effort med/Impact high/Retention + referral flywheel in one feature

04

Nutrition gamification + workout unification

In progress

Beats siloed tracking that fragments motivation

People log nutrition for a few days and then fall off. The data is there, the features work -- the missing piece is the motivational layer. Streaks, badges, and momentum already drive workout consistency. Right now, none of that exists for nutrition. Step one is applying the same gamification logic to nutrition that already works for training: daily logging streaks for macros, water, and steps; badges for hitting calorie targets, protein targets, and "Locked In" days (both at once); a Strength Series macro challenge. Build the logging habit first, then reward it the same way workouts get rewarded. Once both sides have that motivational infrastructure, the second layer becomes natural: make training and nutrition aware of each other. When you hit your protein target, the coach knows. When you miss it before a heavy session, it says something. Squad visibility into fueling (not calorie count, just "fueled / not fueled") creates light social accountability without oversharing. The identity that sticks is: "I'm someone who trains and fuels right" -- stickier than either half alone.

Effort med/Impact high/Nutrition gamification is Q3/Q4 roadmap

05

Temptation bundling

Half-built

Beats no immediate reward for the workout

Milkman's signature move: enjoy a binge-worthy "want" only while doing the "should." For a lot of people that want is an audiobook or a podcast. The idea is simple: you only get to listen to your current audiobook, your favorite podcast, or a serialized audio drama while you're actually working out. The chapter ends when the workout ends. Ladder owns the audio channel and has volume ducking, but the coach voiceover occupies that same channel, so a full audiobook swap gets complicated. The realistic version: target the sessions where the channel is free. Coach-audio-off mode, steady-state cardio sections, rest blocks, Flex workouts. An "only while you train" audio layer lives there without fighting the coach. A serialized podcast or audio drama partnership would be the premium version. Where audio is genuinely contested, the bundle could be non-audio instead (a show that only unlocks post-workout, for example).

Effort med/Impact high/Owns the channel

06

Flexible streaks

Recalibrate

Beats the "what-the-hell" collapse

Streaks are powerful, but rigid streaks backfire. Milkman's research shows that missing once triggers the "what-the-hell" effect: you feel like you've blown it, so you stop entirely. The insight: the problem isn't the streak mechanic, it's the brittleness. A Ladder streak requires 3 workouts per week. Miss that threshold and the streak resets to zero. Streak freezes exist as a buffer, but earning them feels harder than it should -- which manufactures the exact fragility the lever is supposed to prevent. The fix: make freezes easier to unlock, build in a small standing reserve so members always have one in pocket, and soften the psychological loss when a streak breaks rather than amplifying it. Flexible consistency outlasts rigid consistency. This is a mechanic re-tune, not a new build, and the retention impact is outsized.

Effort low/Impact high/Sharpest cheap win

07

Commitment devices

New build

Beats present bias & impulsivity

Current commitment mechanics are soft (badges, streaks). Prototype harder ones that let a member bind their future self: a self-set weekly target, optional cash/charity stakes, or a public pledge to the squad. The annual plan is itself a commitment device - lean into that psychology in messaging, not just pricing.

Effort med/Impact med-high

08

Advice-giving

New build - the sleeper

Beats low confidence, the month-two wobble

Counterintuitive finding: giving advice lifts the giver's own motivation more than receiving it - strongest for people who feel like they're struggling. Ladder's cheers and selfie wall are all receiving-side. Flip it: let mid-journey members give a tip to newcomers, add lightweight mentor roles inside teams, surface a "what helped you show up this week?" prompt to brand-new members. Turns a wavering user into someone with a reason to stay.

Effort low/Impact med-high/Almost free

C · Resurrection

Bring them back

09

Fresh starts

New build

Beats baggage from past failure

Temporal landmarks create a clean-slate feeling that resets motivation. So reframe every lapse around one - "new week starts Monday," first of the month, a birthday, January - and time win-back pushes to those moments. Never lead a re-engagement message with the broken streak; that shames and repels. Frame re-entry as a fresh start, not a resumption of failure.

Effort low/Impact med/High on churned segment

Three product realities folded in
On the audio play

Coach voice owns the channel

A straight audiobook swap fights the coach voiceover. So the bundle lives where the channel is free: coach-audio-off sessions and ducked rest blocks, not the whole catalog. Treat it as a mode, not a default.

On streak freezes

The freeze economy is inverted

A streak requires 3 workouts per week. Freezes exist, but they're harder to earn than they should be, which manufactures the exact fragility the lever is meant to prevent. The fix isn't a new feature; it's re-tuning earn rate, reserve size, and loss penalty so members always have a buffer in pocket.

On nutrition gamification

Build the habit first, then connect it

The sequence matters. Gamify nutrition logging first (streaks, badges, Locked In days) so members build the habit. Then make training and nutrition aware of each other. The unification layer only works once both sides have momentum.

The climb, sequenced

Where I'd start

These are loose thoughts on sequencing, not a prescriptive roadmap. The right order obviously depends on where the 2H product roadmap is already headed and what's in flight. Use this as a thinking frame, not a ranked backlog.

Ship first - low effort, clear impact

1
Recalibrate the streak-freeze economy.Re-tuning, not a build. Directly stops the "what-the-hell" churn already baked into the streak mechanic. Make freezes easier to earn and keep one in reserve.
2
Nutrition gamification -- streaks and badges for logging.Already on the Q3/Q4 roadmap. Builds the logging habit with the same motivational layer that drives workout consistency. Foundation for everything else in the nutrition stack.
3
Squads Phase 1 -- connect with people you know.Retention + referral flywheel in one feature. Descriptive norms (teammate activity counts) are a low-lift layer on top. Phase 2 opens to tribes.

Bigger bets - sequence after

4
Implementation intentions -- scheduling, calendar sync, recurring.The current scheduling feature barely exists. A proper implementation intentions layer (GCal sync, recurring schedule, cue-based reminders) would be a significant upgrade to activation.
5
Temptation-bundling audio layer.Ladder uniquely owns the audio channel. "Only while you train" podcast or audiobook access is defensible and hard to copy.
6
Re-skin win-backs as fresh starts.Mostly copy + timing. Stops shame-framing from repelling the lapsed segment. High impact on the churned cohort specifically.