Habit-Formation Model for Team Behavior Change
Drive team adoption (e.g. of AI) with behavioral psychology, not education: consistency, low friction, and a powerful reward loop.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 90%
A model for changing team behavior that rejects the education/enforcement approach (teach, assess, hold accountable) in favor of behavioral psychology. Three levers: build consistency by making the behavior tiny and daily; strip out friction so the behavior is effortless to start; and design a reward loop that is powerful, immediate, and emotional so people feel great every time they do it. Applied by Gridley to drive AI adoption on her team and mirrored in Whoop's product design.
Origin
Developed by Hilary Gridley at Whoop, drawing on her interest in habit formation and reward loops and on Whoop's own product mechanics (e.g. the recovery score).
Core principles
- 01Behavior change works better as behavioral psychology than as an education model of teach-assess-enforce.
- 02Consistency requires starting small — tasks that take no more than a minute or two.
- 03The last thing anyone wants when under deadline is more friction, so practice on low-stakes tasks unrelated to real work.
- 04The reward loop needs to be powerful, immediate, and emotional — the person should feel 'like a million bucks.'
- 05Reward the behaviors you actually want to encourage; inadvertently rewarding 'stayed up till 2am to ship it' breeds that behavior.
How to run it
- 1
Engineer consistency with tiny daily reps
Get the person doing the behavior every single day, starting with tasks that take one or two minutes. Structure it as habit formation, not a curriculum.
Pro tip Gridley's '30 days of GPT' gives one small AI task per day; she says nobody who completes it comes out less than a hundred times more confident.
- 2
Remove all friction
Have people practice on fun, low-stakes tasks with no external work pressure — planning a vacation, generating meeting talking points from a calendar — so there's no frustration of a bad result on real work.
Pro tip Avoid starting with the person's actual job; work is hard and layering an unfamiliar tool on a deadline creates friction that kills the habit.
Watch out If the first experiences are frustrating ('I'm not good at prompting and this didn't work'), the habit dies before it forms.
- 3
Design a powerful, immediate, emotional reward loop
Ensure that when the person does the behavior, they get an immediate emotional payoff. Build custom GPTs pre-loaded with a good prompt so a novice uploads a doc and instantly gets a useful result — joy without the despair of bad prompting.
Pro tip Give public shout-outs and let people demo their wins in team meetings — make them feel like a million bucks for doing the behavior.
- 4
Reward only the behaviors you want more of
Be deliberate about which behaviors your praise reinforces. Publicly rewarding overwork trains overwork; instead reward the behaviors (like self-care or smart AI use) you actually want to see repeated.
Pro tip Catch people in already-good moods (e.g. after a light team moment) to deliver the shout-out for maximum emotional impact.
Watch out Praising 'they stayed up till 2am but got it done' inadvertently makes staying up till 2am the modeled behavior.
In the wild
Whoop gives a red/yellow/green recovery score each morning. People who drink quickly learn that drinking produces a red recovery. Nobody was unaware alcohol hurt their sleep, but seeing the red score creates a profound, immediate emotional impact, and the green score feels great. Members report getting a handle on years-long drinking problems only after getting on Whoop — the immediate emotional reward loop overrode the behavior in a way that information alone never had.
→ An immediate, emotional reward loop changed a hard behavior that pure information had failed to change for years.
Common mistakes
Using an education/enforcement model
Teaching, measuring, and enforcing a behavior ignores how habits actually form and produces weaker adoption than a consistency-plus-reward approach.
Under-investing in the reward loop
Gridley's number-one note to anyone designing for behavior change: you're not thinking enough about the reward loop — without a powerful, immediate, emotional payoff, the behavior won't stick.
Is it for you?
Best for
Managers driving adoption of a new practice or tool (like AI) across a team, and anyone designing products or systems meant to change human behavior.
Not ideal for
Mandatory compliance behaviors where enforcement and accountability are genuinely required and intrinsic reward can't be engineered.
From the transcript
“the reward loop needs to be powerful. It needs to be immediate and it needs to be emotional so that when this person does the…”
“I think a mistake a lot of people make when they start thinking about how to drive adoption is they're like, "Oh, we have to…”
“you are not thinking enough about the reward loop”
From the episode
How to build a team that can “take a punch”: A playbook for building resilient, high-performing teams
Hilary Gridley (Head of Core Product, Whoop)