Applying behavioral science to improve product design and conversion

Executive overview

Most product teams build features based on what users say they want — but users' stated preferences rarely predict their actual behavior. Behavioral economics reveals that people act emotionally, follow defaults, and prioritize the present over the future, in predictable ways that can be designed around.

Kristen Berman of Irrational Labs uses a 3B framework — Behavior, Barriers, Benefits — to help product teams diagnose why users aren't doing what the product needs them to do, then design targeted interventions.

The core insight: behavior change requires redesigning the environment, not just setting goals — and the most effective interventions reduce cognitive friction and deliver immediate, concrete benefits.

The 3B framework

• Behavior: Define the single, uncomfortably specific action you want users to take (e.g., "within 7 days, do two 10-minute workouts with two different instructors") — not a vague outcome like "engagement"
• If teammates aren't arguing about which behavior to target, you're not being specific enough
• Wrong answer: "log in" — it's what happens after login that matters
• Barriers: Two types — logistical (forms, wait times, steps) and cognitive (uncertainty aversion, status quo effect, information aversion)
• Benefits: Users are present-biased; you must give them a reason to act today, not just explain the long-term value
• Completion bias, social desirability, and social norms are powerful immediate motivators

Behavioral diagnosis

• Map every step required to reach the target behavior — including steps before users hit the product
• Overlay the psychology driving (or blocking) users at each step
• Creates "light bulb" moments for teams: makes it obvious which interventions are worth pursuing
• Functions like a journey map, but focused on what users actually do rather than what they say they will

Common high-leverage interventions

• Reduce choice: Recommending one provider with a reason (OneMedical) increased appointment bookings 20% during onboarding
• Ask engaging questions during sign-up: Gets users thinking about benefits; Trunk Club reported 133% conversion lift; a quiz on TitoKare's site lifted purchase rate from 37% to 53%
• Add friction to reduce unwanted behavior: A label plus a share confirmation pop-up on TikTok reduced misinformation shares by 24%
• Defaults: Automatic enrollment is why Americans have retirement savings — make the right behavior the path of least resistance
• Rules of thumb over decisions: Replace in-the-moment trade-offs with pre-committed heuristics (e.g., "I don't take Lyft on weekdays")
• Resume where users left off: Returning users to their drop-off point (not the start) increases completion — Wealthfront identified this as a key insight
• Deadlines: Adding a deadline to Kiva's borrower onboarding increased conversion; deadlines help people prioritize what they already want to do

Right for wrong

• The "right" behavior is achieved using a motivator that isn't the "real" reason — giving people an immediate hook
• Vaccine uptake increased with donuts and lottery tickets; voter turnout rises with pizza at polling lines
• Peloton streaks and instructor shout-outs motivate workouts more immediately than health goals
• Use with caution: the end outcome must genuinely benefit the user

The dark side: incentives determine outcomes

• What your team measures shapes how behavioral tools get used
• Setting incentives on the behavior (not just retention or MAU) aligns team goals with customer outcomes
• Extending incentive time horizons (annual vs. quarterly) reduces short-term manipulation pressure
• Lending Club example: payment tied to a conversion bump pushed toward predatory tactics — incentive design matters as much as behavioral design

Case study: TikTok misinformation

• Target behavior: reduce shares of misinformation videos (not likes, comments, or general engagement)
• Intervention: unverified content label + "are you sure?" confirmation on share
• Result: 24% reduction in shares
• Process: literature review → 30+ hypotheses → 5 pop-up variants tested with 1,000+ users on Prolific → two conditions launched in product

Case study: OneMedical onboarding

• Problem: users sign up then revert to old habits when they get sick months later
• Target behavior: book a doctor's appointment during onboarding
• Barriers: choosing a provider, finding an available time, uncertainty about what OneMedical covers
• Intervention: brief health quiz → one recommended provider with rationale → limited appointment times (virtual, tomorrow)
• Result: 20% increase in appointment bookings during onboarding

How to start without a behavioral science team

• Run a workshop to align on one uncomfortably specific target behavior
• Conduct a behavioral diagnosis: screenshot every step, attach the blocking psychology to each
• Search Google Scholar before talking to users — find what's already been studied
• Test multiple conditions against each other (not just one idea in isolation); relative comparison predicts in-market results better than absolute ratings
• Use the 3B framework as a lightweight audit of any existing flow