Behaviour design: find the field that blocks

Why the classic techniques work, and how to match the technique to the barrier

Most behaviour design is a list of tricks: social proof, defaults, scarcity, and so on. But there is one mechanism underneath all of them, and once you can see it they turn into a single picture. An action happens when it wins an internal race. It fails to happen when one of four fields makes it too costly: safety, meaning, ability, or effort. So the skill is not knowing twenty tricks. It is working out which field is blocking, and then using the technique that works on that field.

The one idea

Every action runs as a race between routes, and the first usable route wins. A technique works by doing something to that race: making the wanted route cheaper, making the unwanted one more expensive, or shifting which field decides.

The four fields are four ways an action can be too costly:

• Safety — does it feel safe to do?
• Meaning — can the person see why?
• Ability — do they know how, and can they?
• Effort — is it easy enough?

Diagnosis comes first

The most common mistake is reaching for your favourite trick without asking which field is actually in the way. An initiative that assumes an ability problem and delivers a course, when the real barrier is safety, solves the wrong friction. It is the same point as the compliance page: more information only helps if knowledge was what was missing.

So start by finding the barrier. Do people hesitate because it feels unsafe? Ignore it because they cannot see the point? Want to but do not know how? Or know exactly what to do, but it is too much hassle? Once you can name the field, it points to the technique itself.

The techniques, sorted by field

Safety: make it safe to act

• Social proof — "others did it and it went fine" lowers the uncertainty. It works because we forward-model other people as a source of what is safe; the weight we give the source is a reliability estimate, not a contradiction.
• Authority and expertise — a credible sender lowers the doubt about what the right thing to do is.
• Guarantees and easy returns — make the action safe to undo, so it does not feel like a leap.
• Familiarity — the known reads as low-risk, so hesitation drops.

Meaning: give the action a reason

• Identity and consistency — when the action fits who you are, or something you already said, the route wins more easily.
• A real "why" — a reason, not just an instruction, strengthens the route toward the wanted action.
• Reciprocity and fairness — social and moral obligations pull, because they run as friction between people.
• Stories and concrete examples — a lived example is a shorter race than pure theory, so it lands faster.

Ability: make the next step clear and doable

• Small steps — a step small enough to be taken makes the route wider; a big step that never gets taken does nothing.
• Help right when it is needed — a short checklist in the moment beats a thorough course given in advance, because it triggers one specific route exactly when it is needed.
• Salience and feedback — make the right route visible in the moment; that steers which race even opens.

Effort: lower the price of the right thing

• Defaults and fewer steps — the pre-chosen option wins because it is cheapest; remove a step and more people reach the end.
• Clear the obstacles — the most overlooked improvement is removing hassle, not adding motivation.

Timing: the first thing and the scarce thing

• Anchoring — the first number or word shapes the rest of the race.
• Scarcity and loss — a loss weighs more than an equivalent gain, so "this is going away soon" raises the price of not acting now.

Friction the other way: designing against the unwanted

Friction is not always the enemy. To make a wrong or irreversible action rarer, make it more expensive on purpose: a confirmation step before deleting, a small obstacle before an impulse buy, an extra click before something is sent. The rule is that friction against the unwanted should feel like care, not bureaucracy.

Why this is not just another list of tricks

Most of the techniques are not new. What is new is that they get one shared explanation, and that each one points to a field, so you can choose by what is actually blocking instead of guessing. And several of the techniques turn up again in language models, where you can measure them directly: anchoring, fairness, and that pushing too hard backfires (see What language models reveal about humans).