Framework ⚡ Medium

Behavioural Design

People do not behave rationally. Design for how they actually decide.

Behaviour Change Habit Formation Onboarding Design Financial Products Health Apps Conversion Optimisation

01 — TL;DR

Two sentences.

Behavioural Design is the application of behavioural economics and cognitive psychology to product decisions. It represents a foundational departure from the rational actor model: users will not do what is best for them simply because the correct action is available. They will do what is easiest, most salient, and least effortful — and Behavioural Design works by reshaping the choice environment to close the gap between intention and action.

The field draws on Daniel Kahneman and Amos Tversky's prospect theory, Richard Thaler and Cass Sunstein's nudge theory, and BJ Fogg's behaviour model. Together these establish a constraint and an opportunity: users will not behave as intended just because the correct action is present, but they will behave differently when the context around the choice changes. The most effective behavioural interventions feel invisible — they make the desired action the path of least resistance rather than the product of deliberation.

02 — When to Use

Apply this when…

→ You are building a behaviour gap product — one where the value depends on users doing something they already want to do but consistently fail to follow through on (saving money, exercising, learning a language)

→ Feature adoption is low despite expressed interest — users say they want the feature, analytics shows they never use it, and the gap is not a usability problem

→ Onboarding completion is low and drop-off happens at action steps, not comprehension steps — users understand what to do but do not do it

→ You are designing defaults — any screen where a pre-selected option will be accepted by the majority of users without change

→ Your user population has a known intention-action gap — health apps, financial products, sustainability tools, education platforms

→ You are designing notification and prompt systems where timing, framing, and frequency determine whether users act or disengage

When NOT to apply it

Skip it for high-stakes deliberate decisions where users need full information and reflection time — mortgage applications, medical consent, legal agreements. Skip it when the behaviour change serves the product rather than the user — that is persuasion architecture, not Behavioural Design. And be cautious in regulated contexts where nudging may conflict with informed consent requirements or fair treatment obligations.

03 — How It Works

The mechanism

Behavioural Design works by redesigning the choice context rather than the choice itself. Instead of informing users and hoping they act rationally, it changes what is easiest, most visible, and most immediate — shaping the environment so the desired behaviour becomes the path of least resistance.

System 1, System 2, and the gap between them

Daniel Kahneman's dual-process theory distinguishes between System 1 (fast, automatic, effortless) and System 2 (slow, deliberate, costly). Most product decisions are designed for System 2 — they assume users will read, compare, and choose optimally. In practice, the majority of user behaviour runs on System 1. The intention-action gap exists because users form intentions using System 2 but must execute them under System 1 conditions — distracted, time-poor, and defaulting to the easiest option. Behavioural Design closes this gap by making the right action the System 1 action.

Six core tools

Defaults: pre-select the option that serves the user's stated goal — most users will not change it. Friction reduction: remove every unnecessary step between intention and completion. Social norms: show users what others like them have done. Commitment devices: let users bind their future selves to present intentions. Implementation intentions: prompt users to specify when, where, and how they will act. Loss framing: show what the user stands to lose by not acting, which is psychologically more powerful than showing what they gain.

The best behavioural design is invisible

When Behavioural Design works well, the user does not notice it — the product simply feels easier and the right action feels natural. This is the ethical distinction from manipulation: Behavioural Design aligns the choice architecture with the user's own goals. The user wanted to save money; the product made saving the default. The user wanted to exercise; the product removed three steps from the workout start flow. Invisible design in service of the user's goal is good design. Invisible design against the user's goal is a dark pattern.

Measurement: behaviour, not satisfaction

You measure Behavioural Design by behaviour completion rates, not user satisfaction scores. The key metrics are: the intention-action gap (what percentage of users who express intent actually complete the behaviour), time-to-action (how quickly users move from intention to completion), and long-term retention (whether the behaviour persists after the initial intervention period). If satisfaction is high but behaviour completion is low, the product is informing users without enabling them.

Behavioural Design vs dark patterns

Behavioural Design and dark patterns use exactly the same techniques — defaults, friction manipulation, loss framing, social proof. The difference is ethical orientation, not method. Behavioural Design uses these tools to help users do what they already said they wanted to do. Dark patterns use them to make users do what the business wants against their own interests. Same toolkit, opposite direction. If you cannot articulate whose goal the intervention serves, you are probably building the wrong thing.

04 — Real Example

Nest thermostat — learning defaults and social norm framing

The traditional programmable thermostat is a textbook case of the intention-action gap. Users want to save energy and money — surveys consistently show this. But programming a thermostat requires deliberate System 2 effort: reading a manual, setting schedules, adjusting for weekends. The result: the majority of programmable thermostats are never actually programmed. The feature exists, the intention exists, the behaviour does not.

Nest solved this with Behavioural Design rather than better UI. Instead of making the programming interface easier, they eliminated it. The Nest thermostat observes what temperature users set manually over the first week, then auto-generates a schedule from that behaviour — learning defaults. Users do not need to programme anything; their existing behaviour becomes the programme. The Energy History screen then applies social norm framing ("You used 12% less energy than similar homes this month") and loss framing ("Turning up 2 degrees would cost you an extra $23 this month"). The combination of defaults, social norms, and loss framing closed the intention-action gap that decades of better thermostat interfaces had failed to address.

Nest · Smart Home · 2011–present

Two approaches to energy saving

Defaults closed the intention-action gap

05 — In the Wild

Test yourself & see real examples

No examples yet — be the first.

Found a product that genuinely helps users do what they said they wanted to do? Or one that uses behavioural techniques against user goals? Submit what you found. Every approved example gets attributed to you.

✓ Reviewed before publishing ✓ Your name on every example you submit ✓ Violation or fix — both welcome

Seen Behavioural Design applied well or misused? Help grow the evidence base.

06 — Common Mistakes

Where teams go wrong

✗ Confusing Behavioural Design with persuasion architecture. Persuasion architecture optimises for business outcomes — conversions, upsells, engagement minutes. Behavioural Design optimises for user outcomes — helping people do what they already said they wanted to do. The techniques overlap almost entirely, but the ethical orientation is opposite. If you cannot name the user goal your intervention serves, you are doing persuasion, not Behavioural Design.

✗ Applying tools without understanding the specific barrier. Teams learn about defaults and apply them everywhere, or learn about loss framing and use it on every notification. Each behavioural tool addresses a specific type of gap: defaults address effort barriers, implementation intentions address planning failures, social norms address uncertainty about appropriateness. Applying the wrong tool to the wrong barrier produces no effect and generates false evidence that "behavioural design does not work here."

✗ Setting defaults that serve the product, not the user. The most powerful behavioural tool is the default — and the most commonly abused. Pre-selecting the premium plan, opting users into marketing emails, defaulting to public profiles — these are defaults designed for business metrics, not user goals. The ethical test is simple: if users understood what the default was doing, would they thank you or resent you? Defaults should make the user's own expressed preference the path of least resistance.

✗ Using engagement as a proxy for wellbeing. A fitness app that sends six push notifications a day may increase daily opens, but if users are not actually exercising more, the engagement metric is measuring the intervention's annoyance surface, not its effectiveness. Behavioural Design measures success by the behaviour it was designed to support, not by the engagement it generates along the way. High engagement with low behaviour completion is a warning sign, not a success metric.

07 — Variations & Related Principles

Connected ideas

Behavioural Design does not operate in isolation — it draws on and complements several related frameworks and strategies. The principles below either explain why specific behavioural tools work, shape how users experience interventions, or provide the testing infrastructure to validate which interventions actually close the gap.

The most important pairing is Behavioural Design with a clear definition of the user's own goal. Without knowing what the user is trying to achieve — in their terms, not yours — every behavioural intervention risks sliding from nudge into manipulation. Cognitive Load Theory explains the mechanism behind friction reduction: every unnecessary step imposes extraneous load that competes with the user's intention. Framing provides the theoretical basis for why loss framing outperforms gain framing in almost every behavioural context. And Lean UX provides the experimental infrastructure to test which specific intervention closes the gap for your specific population.

08 — 10-Min Exercise

Run it right now

⏱ 10 minutes · Team · No prep

The Intention-Action Gap Audit

Pick one behaviour your product is designed to support — one action users should take that they currently do not take often enough.

1. Estimate the intention-action gap: what percentage of users who express intent (sign up, start onboarding, add to cart, set a goal) actually complete the target behaviour? If you do not have data, use your best estimate — the number is almost always lower than the team believes.

2. For each of the six core tools — defaults, friction reduction, social norms, commitment devices, implementation intentions, loss framing — write one sentence describing how it could be applied to this specific behaviour gap. Not every tool will be relevant; skip any that do not fit.

3. For the two most promising interventions, write a testable hypothesis: "If we [specific change], we expect [specific behaviour metric] to increase by [estimated percentage] within [timeframe]."

4. Identify the smallest possible experiment you could run this sprint to test the stronger of the two hypotheses. What is the minimum change to the choice architecture that would generate measurable signal?

10 minutes