LLenny's Podcast
← All episodes
Kristen Berman (Irrational Labs)02 October 2022

Using behavioral science to improve your product

8Frameworks
13Insights

Frameworks in this episode

Influence4 steps

Benefit-Priming Questions (Good Friction)

Add a step to your signup flow — a question that makes users think about the benefit — and conversion goes up.

Read →
Strategy5 steps

Evidence-First Intervention Design

Literature review, then hypothesis, then 30 variants, then relative testing — never show users one design.

Read →
Influence4 steps

Friction Injection for Behavior Suppression

To make people do less of something, add logistical AND cognitive friction at the moment of the hot decision.

Read →
Leadership4 steps

Incentive Alignment Against the Dark Side

We are what we measure. Set the team's incentive on the behavior, and lengthen its time horizon.

Read →
Influence4 steps

Right for Wrong

Get people to do the thing they already want to do — by giving them a shallower, more immediate reason.

Read →
Innovation5 steps

The Behavioral Diagnosis

A journey map on steroids: every micro-step to the behavior, with the blocking psychology named at each one.

Read →
Innovation4 steps

The Easiness Ladder: Default, Then Rule of Thumb, Then Decision

First line of defence is always a default. If you can't default it, give people a heuristic — never a calculation.

Read →
Innovation4 steps

The Three B's (Behavior, Barriers, Benefits)

Pick an uncomfortably specific behavior, strip its barriers, then engineer an immediate benefit.

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster11:30

When Adding Friction Increases Conversion

The rule everyone lives by is that reducing signup friction increases conversion, but Kristen shows one case where the opposite holds. Asking a user a question inserts an idea into their head, getting them to engage with the product's benefits, which raises motivation to push through later friction. Trunk Club reported this approach lifted conversion by 133 percent.

  • Asking a question during signup makes users think about the benefits, raising motivation
  • Trunk Club said question-based signup increased conversion by 133 percent
  • A quiz for a health device: 53% of quiz-completers purchased vs 37% who didn't complete
  • Use easy dropdown or multiple-choice questions, never hard open-text fields

“one example one time where this is just not true that you can actually increase friction and increase conversion”

Kristen Berman · 12:00

“when you ask a question you can insert an idea into someone's head you can get them thinking about something different”

Kristen Berman · 12:00
#conversion#signup-flow#friction#product

Hot Take· 2

Hot Take45:00

Deadlines Are A Gift

Nonprofits often resist adding deadlines because they feel harmful, but Kristen argues deadlines actually help people prioritize and follow through. Across multiple companies, adding a deadline increased the number of people who completed the desired behavior. Adding a deadline to Kiva's borrower signup flow increased their conversion.

  • Nonprofits resisted deadlines, fearing they'd be too hurtful
  • Deadlines help people prioritize a behavior they already want to do
  • Adding deadlines increased completion across multiple companies
  • A deadline on Kiva's borrower signup flow increased conversion

“sometimes we say deadlines are a gift right you're just helping people kind of prioritize this”

Kristen Berman · 45:00

“when you add deadlines and we've done this in multiple companies where you add a deadline more people do the thing”

Kristen Berman · 45:30
#deadlines#conversion#nonprofits#behavior-change
Hot Take54:00

Interviews Don't Predict Job Performance

Asked for a favorite interview question, Kristen first pushes back on interviews themselves: there's little to no evidence they predict on-the-job performance. Her team relies on skill assessments and trials to predict performance, and uses interviews only for affinity and culture fit. Her favorite culture-fit question asks what personality trait defines you that's both a strength and something to improve.

  • Interviews have little to no evidence for predicting job performance
  • Use skill assessments and trials to predict actual performance
  • Interviews are good for affinity and culture fit
  • Favorite question: a personality trait that is both your strength and your weakness

“interviews actually we know don't predict job performance so very little to no evidence and say interviews will predict how I perform on the job”

Kristen Berman · 54:00
#hiring#interviews#job-performance

Explainer· 2

Explainer06:00

What Behavioral Economics Actually Is

Kristen contrasts classical economics, which assumes people are rational and emotionless, with behavioral economics, which fuses psychology and economics. The key insight is that people make emotional, biased decisions, but they do so in predictable ways, which means those behaviors can be understood and changed.

  • Classical economics assumes people are rational and weigh pros and cons with no emotion
  • Behavioral economics combines psychology and economics
  • People are present-biased, overweight their present selves, and follow social norms
  • Irrational behavior happens in predictable ways, so it can be designed for and changed

“economics says look people are rational we make decisions with no emotion we use lots of computational energy weigh the pros and cons I mean…”

Kristen Berman · 06:00

“the good news is that we do these things in predictable ways and once you understand how and why people behave you can start to…”

Kristen Berman · 06:30
#behavioral-economics#psychology#product-design
Explainer42:30

'Right For Wrong': Giving People The Wrong Reason To Do The Right Thing

Kristen explains 'right for wrong reason' — getting people to do a behavior they want to do by supplying an immediate, unrelated motivator. People got vaccines for donuts and lottery tickets rather than to protect neighbors, and one of the biggest voting interventions is simply putting pizza in the poll line. The caveat: the end goal must genuinely be good.

  • People do a good behavior for an immediate, unrelated reason
  • Vaccine uptake rose with donuts, free marijuana, and Ohio lottery tickets
  • Pizza in the poll line is one of the bigger voting interventions
  • Ethical guardrail: the end goal itself must be good for the user

“one of the bigger voting interventions is just having pizza show up at the pole line right you're motivating people to go vote not because…”

Kristen Berman · 43:00
#motivation#immediate-benefits#behavior-change#ethics

Story· 4

Story07:00

The Most-Requested Feature That Changed Nothing

A popular fintech app's most-requested feature was budgeting, so they nearly built it on demand. Irrational Labs convinced them to run a 10,000-person experiment with a control and two budgeting variations instead. The result: zero change in average spend and zero change in spend variability, which shocked the team but not the behavioral scientists.

  • Budgeting was the fintech app's most-requested feature across support forums and interviews
  • The experiment used 10,000 people across a control and two budgeting variations
  • Result was zero change in average spend and zero change in spend variability
  • Mapping every step of a behavior reveals when a feature can't pull the weight

“and what happened nothing so they were shocked the team was shocked”

Kristen Berman · 08:00

“zero change in average spend zero change in spend variability”

Kristen Berman · 08:00
#fintech#experimentation#budgeting#product
Story25:00

How She Almost Became A Predatory Lender

Lending Club paid Kristen a large bonus only if she hit a five-point conversion bump for borrowers. Chasing that incentive, she started suggesting interventions the legal team had to shut down, becoming a predatory lender even while simultaneously running a Duke lab focused on the financial health of low-income Americans. The lesson: outcomes are driven by the incentives you set, not the tactics.

  • The contract paid a large bonus for a five-point conversion bump and zero otherwise
  • The incentive structure pushed her toward predatory suggestions the legal team blocked
  • She was simultaneously leading Duke's Common Sense Lab on low-income financial health
  • 'We are what we measure' — incentives, not tactics, determine how tools get used

“over the course of working with them I became a predatory lender I was suggesting things that the legal team was like no Kristen of…”

Kristen Berman · 25:30

“I like to say kind of we are what we measure and it really matters what you measure”

Kristen Berman · 25:00
#incentives#ethics#fintech#conversion
Story29:30

How TikTok Cut Misinformation Shares By 24

TikTok asked Irrational Labs to reduce the spread of misinformation, and the team narrowed the target behavior to decreasing shares specifically. Because you want people to do something less, they added barriers: an unverified-information label plus an 'are you sure' pop-up after the share button. Citing an elevator study where a slower door pushed people to the stairs, the interventions together reduced shares by 24.

  • Target behavior was narrowed to decreasing shares, not likes or comments
  • To reduce a behavior you add barriers instead of removing them
  • An elevator study made the door close 16 seconds longer so people took the stairs
  • A label plus an 'are you sure' pop-up added logistical and cognitive friction, cutting shares by 24

“make the elevator door close 16 seconds longer so now you're getting the elevator you're annoyed it's harder to use the elevator and people went…”

Kristen Berman · 30:30

“the label on the video and the pop-up reduce shares by 24”

Kristen Berman · 31:30
#tiktok#misinformation#friction#case-study
Story35:30

One Medical: Book The First Appointment During Onboarding

For One Medical, the team defined the key behavior as booking a doctor's appointment right after signup, because users who don't set the mental model early will just call their old doctor when they get sick months later. They reduced choice by recommending a single provider with a reason, and offered limited, near-term virtual appointment slots. The onboarding intervention increased bookings by 20.

  • Key behavior: book an appointment immediately after signing up
  • Onboarding must set the mental model so users remember the app when sick later
  • Reduce choice: recommend one provider and explain why
  • Limited, near-term virtual appointment slots increased bookings by 20

“instead of having lots of choice decrease the choice recommend them one provider told them why and then recommended appointment times”

Kristen Berman · 37:30

“the onboarding intervention increased bookings by 20”

Kristen Berman · 38:00
#one-medical#onboarding#engagement#case-study

Tool· 1

Tool52:00

Kristen's Behavioral Science Book Picks

In the lightning round Kristen recommends starter books for anyone getting into behavioral science. She names Predictably Irrational and Influence as great entry points, and flags an under-talked-about pick: Robert Frank's Darwin Economy, on status, social norms, and how they drive behavior.

  • Predictably Irrational — great starter book for the field
  • Influence — another powerful starter book that can change your career
  • Robert Frank's Darwin Economy — under-discussed, covers status and social norms

“I would be remiss not to say particularly irrational I think this is a great starter book for people getting interested in the field influence…”

Kristen Berman · 52:00
#books#recommendations#behavioral-science

Takeaway· 3

Takeaway09:30

Defaults And Rules Of Thumb Beat Willpower

Instead of asking users to do hard cognitive work like budgeting, the first line of defense is to make the behavior easier. In personal finance the best answer is usually a default, like automatic 401K enrollment. When defaults aren't feasible, rules of thumb make decisions cognitively easier so people stay adherent.

  • First question in any problem: how do you make it easier for someone to do?
  • Defaults are why 401K auto-enrollment created US retirement savings
  • Rules of thumb (e.g. 'no Lyfts on weekdays') remove in-the-moment cost-benefit calculation
  • A heuristic isn't perfect but keeps people adherent

“in the personal finance world really the answer to most things is just default so it worked for automatic enrollment 401K the reason we have…”

Kristen Berman · 09:30
#defaults#personal-finance#behavior-change#heuristics
Takeaway26:30

Set Incentives On The Behavior, Not The Outcome

To keep product work customer-friendly, Kristen offers two tips for setting team incentives. First, tie the incentive to the specific behavior rather than raw active use or retention, since the behavior aligns better with customer outcomes. Second, lengthen the incentive's time horizon so teams optimize for long-term company and consumer interest rather than short-term bumps.

  • Set the KPI incentive on the behavior, which aligns with customer outcomes
  • Still measure active use and retention, but don't incentivize on them alone
  • Longer incentive duration prevents short-term-only optimizations
  • Quarterly targets breed short-termism; annual or longer targets protect the customer

“one would be to set the incentive on the behavior because the behavior is going to align more with the”

Kristen Berman · 26:30

“second would be increasing the duration of the incentive”

Kristen Berman · 27:00
#incentives#kpis#product-management#metrics
Takeaway33:30

Google Scholar Before You Talk To Five Users

Before interviewing a handful of users on a new problem, Kristen advises spending a day searching what has already been studied. Google Scholar and the right domain keywords will surface meta-analyses summarizing prior work. And when testing designs, never show users a single option, because they might love or hate everything; present multiple options and compare relatively.

  • Spend a day Googling and using Google Scholar before user interviews
  • Use the right domain buzzwords to find relevant meta-analyses
  • Never run a UX study showing just one option
  • Always present multiple conditions and compare which drives the target behavior

“go spend a day on the internet Googling to see what else has been done”

Kristen Berman · 33:30

“we never do a study a ux study where we're just showing people one thing”

Kristen Berman · 34:00
#research#user-research#google-scholar#testing