The Three B's (Behavior, Barriers, Benefits)
Pick an uncomfortably specific behavior, strip its barriers, then engineer an immediate benefit.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 95%
Irrational Labs' core behavior-change model for product and marketing teams. Instead of chasing outcomes (engagement, retention, revenue), you name the single user action that produces the outcome, then systematically remove what blocks it and add a reason to do it today. It has been deployed at Google, Microsoft, LinkedIn, TikTok and One Medical.
Origin
Developed by Kristen Berman and the team at Irrational Labs, the behavioral science consultancy she co-founded in 2013 with behavioral economist Dan Ariely. It compresses the academic behavioral-economics literature (present bias, status quo effect, social norms) into three steps a PM can run in a workshop.
Core principles
- 01Companies are good at outcomes and bad at behaviors — an outcome is not something a user can do.
- 02The only wrong answer for a target behavior is 'log in'. What matters is what happens after login.
- 03If your team isn't arguing about the behavior, it isn't specific enough.
- 04Barriers come in two flavors: logistical (form fields, credit cards, wait times) and cognitive (uncertainty aversion, optimism bias, information aversion, status quo).
- 05People are present-biased: the benefit that actually drives the action must be available today, not in a year.
How to run it
- 1
Define the behavior — uncomfortably specifically
Name the exact action, with a count and a time window, that you want the user to take. Berman's example: 'within seven days of somebody starting the app they do two 10-minute workouts with two different instructors.' Run this as a team workshop and force alignment.
Pro tip Test the specificity: run the exercise where everyone on the team separately writes what 'engagement' means for your product, then compare. The answers will diverge, and that divergence is the problem.
Watch out 'Log in', 'engage', 'use the product more' are not behaviors. If you can't define the behavior, you cannot identify the psychologies that affect it — every downstream step collapses.
- 2
Reduce the logistical barriers
List every physical step in the way of the behavior: form fields, credit-card entry, choosing a provider, finding a time slot. Delete, default, or pre-fill as many as possible.
Pro tip If a user drops out of a long flow, land them back exactly where they left off rather than at the beginning — Wealthfront named this as one of their key conversion insights.
- 3
Reduce the cognitive barriers
Name the psychologies operating at each step — uncertainty aversion (people who can't predict an outcome go look for another option or freeze), status quo effect (people default to what they did yesterday), information aversion, optimism bias — and design specifically against them.
Pro tip Status quo effect is the biggest one startups fight: you are asking people to do something different from yesterday, so you must either raise motivation or cut barriers hard enough to overcome inertia.
Watch out Uncertainty doesn't just slow people down — it sends them to a competitor. A Lyft rider unsure when the car arrives opens Uber.
- 4
Engineer an immediate benefit
Identify the real, long-term benefit (a project delivered on time, better health), then find the small psychological payoff that will actually motivate the action right now — completion bias, social desirability, streaks, status. Build that payoff into the feature.
Pro tip Asana's task-logging works partly on completion bias (the satisfying checkbox) and social desirability (a teammate gets notified) — not on the abstract benefit of a well-run project.
Watch out Don't over-invest in confetti and animation. The psychological power is in the sense of completion itself, not the celebration graphic.
In the wild
Asked to 'increase engagement', Irrational Labs first forced a behavior definition: book a doctor's appointment immediately after signup (because six months later, when you're actually sick, you'll call your old doctor out of habit). Barriers: picking a provider, finding a time, not knowing what One Medical is for. Benefit: getting a health question answered immediately. The intervention asked a few health questions during onboarding, recommended one provider with a reason, and offered a short list of appointment times for tomorrow, virtual.
→ Bookings during onboarding increased by 20%.
Berman's illustration of specificity: as a Peloton PM you'd define the target as 'within seven days of starting the app, complete two 10-minute workouts with two different instructors' — even though you'd be happy with one workout and one instructor.
→ The over-specified behavior makes the relevant psychologies visible and gives the team a single, arguable target instead of a vague activation goal.
Common mistakes
Optimizing an outcome instead of a behavior
Teams set KPIs on engagement, retention or revenue — none of which a user can actually perform. Without a named action, you cannot diagnose the psychologies blocking it or design an intervention against them.
Selling the end benefit instead of the today benefit
We want to believe people use the product for its ultimate value. They don't act on it. If the payoff is a year away, present bias means nothing happens today — you have to give a reason to act now.
Is it for you?
Best for
Product managers, growth leads and marketers who have been handed a vague mandate like 'increase engagement' and need to convert it into a testable intervention.
Not ideal for
Zero-to-one product discovery where no user flow exists yet and there is no behavior to diagnose.
From the transcript
“our team basically has created a model of behavior change we call the Three B's”
“companies are really good at outcomes but just not as sharp at picking the behavior when I say Behavior I mean action the thing that…”
“so example if I'm Peloton pm and I'm working on the app I would say something like within seven days of somebody starting the app…”
“second B again is probably pretty obvious and it's very critical it's just barriers so we need to reduce the barriers to doing the behavior…”
“so the third B is benefits so this is where you want to increase the not just the benefits but the immediate benefits of doing…”
“I would say the activation Milestone and the novel thing here is getting really specific about this Behavior so much that if you're not arguing…”
From the episode
Using behavioral science to improve your product
Kristen Berman (Irrational Labs)