011 min
The problem the EAST Framework solves
Picture a product team at a personal-finance app. In a survey, most of its users say they want to save money every month. Yet only a small share of them ever set up a monthly transfer.
The team's first idea is to explain more. They write a longer email about interest and a help page about saving goals. People open the email. The number of transfers does not change.
The EAST Framework starts from a different question. It does not ask what people know. It asks what stands between the plan and the action at the moment of choice. Maybe the transfer form has nine fields. Maybe the reminder arrives late on a Sunday, far from payday. Maybe nobody can see that other people save at all.
Much of daily behaviour comes from habit and quick judgement, not careful thought. So small details of a choice often decide what people do. Researchers call the distance between what people plan and what they do the intention–action gap. EAST is a way to search a service for the details that keep people from acting.
021 min
Where the EAST Framework comes from
The finance team's question was first asked about public services. In 2010, Paul Dolan, David Halpern, Michael Hallsworth and two co-authors wrote MINDSPACE, a report for the Institute for Government. It set out how research on behaviour change could help with policy problems such as crime and obesity.
Officials found MINDSPACE useful but long. The Behavioural Insights Team, a unit that ran field trials for the UK government, wanted something people could remember in a workshop. The Behavioural Insights Team published EAST in 2014, when it had 14 staff and had recently become independent of the UK government. The original authors were Owain Service, Michael Hallsworth, David Halpern and nine colleagues.
In 2024 the team released a revised edition. It added newer studies, examples from more countries, and notes on earlier findings that had not held up. The preface also mentions that the team turned down a joking suggestion from the legal scholar Cass Sunstein to add "Fun" and call it FEAST.
EAST is a principle, not a law. Nobody ran one experiment that proved it. It is a way of sorting hundreds of field studies under four plain words. The report itself says it will not make anyone an expert, and that more complete frameworks exist.
032 min
What the EAST Framework actually says
The EAST Framework turns that search into four questions about any behaviour you want more of.
| Principle | The question it asks | Tactics it covers |
|---|---|---|
| Easy | What effort stands in the way? | Set a good default, cut steps and fields, simplify the message |
| Attractive | Will the person notice it and want it? | Use new or personal information, frame the choice, design rewards and penalties |
| Social | What do other people do and approve of? | State the true norm, use networks, invite reciprocity and commitments to others |
| Timely | Does it arrive when the person can act? | Pick the moment, bring benefits forward, help people make a plan |
Why each EAST principle works
Easy works because effort costs more than it seems. Every extra step is a place where a busy person stops. Defaults matter most, because a default decides what happens when a person does nothing at all. The default effect is the name for this pattern.
Attractive works because attention is limited. People notice what is new, personal or surprising, and skip the rest. Attractive also covers incentives. The shape of a reward or a penalty can matter as much as its size.
Social works because people look at others when they are unsure what to do. Hearing that most people do something makes it seem normal and expected. Reciprocity adds a second reason: people help those who helped them, and those who help others.
Timely works because the same request gets a different response at different moments. A prompt works best just before the action. Benefits that arrive now count for more than benefits that arrive later, a tendency called present bias. Habits are also easier to change when life is already changing, such as after a move or a new job.
Why changing the choice beats explaining it
The four letters are not equal in strength. In 2022, Stephanie Mertens and colleagues combined more than 200 studies of choice architecture, the practice of designing how options are presented. Changes to the structure of a choice, such as defaults, worked better than changes to how the options were described or than reminders of a person's goals. In EAST terms, the strongest tactics usually sit under Easy. The finance team's longer email changed the description. A pre-set monthly transfer would change the structure.
043 min
How the EAST Framework shows up in real services
Each EAST principle comes from trials that changed a real service and counted what people then did.
Social: the tax reminder letter
In the United Kingdom, the tax authority sends letters to people who owe tax. Michael Hallsworth, John List, Robert Metcalfe and Ivo Vlaev tested new wording in these letters with more than 200,000 taxpayers. In the first experiment, about 100,000 people who had declared their income but not paid received one of several versions. One version told the reader that nine in ten people in the UK pay on time, and that they were among the very few who had not. That version raised the number of people paying by 5.1 percent, the largest effect among the wordings tested. A second experiment found that saying what most people do worked better than saying what most people think is right. The authors estimate the tax authority collected £9 million more over 23 days, and the letters cost nothing extra to send. This is a correct use of social proof: the claim was true, and it described people like the reader.
Reciprocity got its own large test. In a trial with one million people, the team tested several messages inviting people to join the UK organ donor register. The best one asked whether the reader would accept a transplant if they needed one, and if so, asked them to help others. The team reports 529,000 extra sign-ups from this message between 2013 and 2017.
Easy: the pension default
The best-known Easy result is a default. The UK changed workplace pensions from opt-in to opt-out, so employees now save unless they choose to leave. Within eight years, about ten million more employees were saving in a workplace pension.
Timely and Attractive: the flu-shot texts
Katherine Milkman and colleagues sent one of 22 text reminders to 689,693 Walmart pharmacy patients in the US to encourage flu shots. On average, the texts raised vaccination by 2.0 percentage points over three months. The best message came as two texts three days apart and told patients a flu shot was waiting for them. It used two EAST principles at once: it arrived more than once, and it presented the shot as already set aside for the person.
Experts could not pick the winning message
Neither experts nor members of the public predicted that this message would win. That is a warning for any team that picks one message in a meeting and ships it untested.
Easy in reverse: the Amazon Prime case
The same principles work in the other direction. If extra effort stops people from acting, a company can add effort on purpose. The US Federal Trade Commission sued Amazon in 2023 over its Prime membership. The agency said some checkout buttons signed people up for Prime without saying so clearly. Cancelling went through a process Amazon staff called Iliad, after the long Greek poem about the siege of Troy. Customers had to confirm on three pages that they wanted to leave. In September 2025, Amazon agreed to a $2.5 billion settlement. Flows like this, which use design against the person, are called dark patterns.
052 min
Applying the EAST Framework
Back at the finance app, the team can now put four questions to its savings flow. The Behavioural Insights Team pairs EAST with a short process, and the steps below follow it.
Target one behaviour you can count.
"Sets up a monthly transfer", not "saves more".
Explore the real flow.
Go through the setup yourself, watch a few users, and count the steps. Find the point where people stop.
Build one change per letter, then choose.
Easy: pre-fill the transfer with a suggested amount the user can change, and cut the form to the fields that matter. Attractive: name the user's own goal in the prompt, such as their trip or their deposit, instead of a general "Start saving". Social: if it is true that most similar users save each month, say so; if it is not true, say nothing. Timely: ask on payday, and let users pick the date of the transfer.
Test against the old version.
Run a randomised trial or an A/B test, and measure transfers, not clicks on the banner.
Scale what works,
and test again when the audience or the moment changes.
Do not expect all four letters to add up. One good change to the default often does more than four changes to wording. Start with the change to structure, and treat the rest as options to test.
Attractive does not mean decorative
A common misreading treats Attractive as a call for nicer visuals. In EAST it means two things: getting the person's attention, and making the action more rewarding than the alternative. The 2024 report gives the example of gold coin stickers on a factory floor in China, which cut the waste left on the floor by more than 20%. The stickers did not make the floor look better. They made the waste harder to ignore.
062 min
When the EAST Framework backfires
A well-chosen EAST tactic can still push behaviour the wrong way, and the finance team should know the three common ways it happens.
A fine that made parents later
In 1998, Uri Gneezy and Aldo Rustichini studied 10 private day-care centres in Haifa, Israel. Some parents came late, so a teacher had to stay after closing. In six centres, the researchers added a fine for parents who arrived more than ten minutes late. The number of late parents went up, not down. The authors gave their paper the title A Fine Is a Price. Once lateness had a price, parents could treat it as a service they paid for, and the guilt that had kept them on time was gone. A penalty belongs under Attractive, and this case shows that its meaning to people matters as much as its size.
Removing the tactic may not undo it
The study had a second result. When the centres removed the fine, lateness stayed at the new, higher level. A team cannot assume it can roll back an incentive and return to where it started.
A default that stuck too well
Defaults have a cost when the pre-set option is a poor fit. Brigitte Madrian and Dennis Shea studied a large US company that switched its 401(k) retirement plan to automatic enrolment. Participation rose sharply. But many new employees kept both the contribution rate and the investment fund the company had pre-set, even though few earlier employees had chosen that pair. The authors traced this to inertia and to employees reading the default as advice from the company. For the finance app, a pre-filled $50 helps a user who would have saved nothing. It may hold back a user who would have chosen $300. Both amounts are illustrative.
Social messages that move people toward the average
Social messages fail in their own way. Telling people the average can move those who were doing better toward it. The 2024 EAST report notes that households using less electricity than their neighbours may use more after a comparison. A norm also fails when few people follow it. If the finance app said "only a few users save", it would make not saving look normal.
Signs that an EAST change is going wrong:
- Clicks on the prompt rise, but the behaviour itself does not.
- Most users stay at the exact default value, whatever their situation.
- Support tickets say leaving is harder than joining.
- A "most people" message cannot be checked against real data.
072 min
Where the evidence behind EAST is contested
Backfires are one problem with EAST. A larger question is how big the effects behind it really are.
The 2024 edition removed an example. The 2014 version cited a car insurance study in which signing at the top of a form made people report more honestly. That study could not be reproduced, and it was later retracted once evidence of data manipulation appeared. The Behavioural Insights Team says its own trials of the idea had not found the same effect.
Smaller effects at scale
Stefano DellaVigna and Elizabeth Linos compared 126 trials run by two US nudge units, covering more than 23 million people, with nudge studies in academic journals. Their working paper came out in 2020 and was published in Econometrica in 2022. The published studies reported an average effect of 8.7 percentage points, while the nudge-unit trials averaged 1.4. The authors found that publication bias could explain the whole gap. Publication bias means journals print striking results more often than results showing no effect.
Does nudging work at all?
Some researchers go further. Maximilian Maier and colleagues re-analysed the Mertens meta-analysis with a method that corrects for publication bias. They concluded that, after the correction, no evidence remained that nudges work on average. They also noted that results varied a lot between studies, so some nudges may still work in some settings.
What this means for a team using EAST
The fair reading is narrower than either side's headline. Average effects are small, and many tactics fail in a given setting. Changes to the structure of a choice, such as defaults, have the strongest record. So EAST works best as a list of ideas to test, not a list of fixes. The 2024 report says the same and pairs EAST with evaluation, ideally randomised trials. The nudge theory debate is the wider version of this argument.
082 min
EAST Framework vs. nearby concepts
Because EAST is a short list, people often mix it up with the models and ideas it draws on.
| Concept | What it is | How it differs from EAST |
|---|---|---|
| Nudge | Thaler and Sunstein's idea that the way options are presented can steer choices without banning any option | Nudging is the general idea. EAST is a working checklist for finding nudges, and it also includes incentives and penalties |
| MINDSPACE | The 2010 Institute for Government report on influences on behaviour, partly by the same authors | A longer list written for senior officials. EAST is the shorter version for teams designing a change |
| COM-B | Susan Michie and colleagues' 2011 model: behaviour needs capability, opportunity and motivation | COM-B diagnoses why a behaviour is missing. EAST goes straight to tactics |
| Fogg Behavior Model | BJ Fogg's model: behaviour happens when motivation, ability and a prompt meet at one moment | Fogg explains one moment of action. EAST lists ways to change the setting around it |
One axis separates EAST from all of these. EAST does not try to explain why people behave as they do. It groups tactics that have worked in field trials, so a team can remember and test them. When a behaviour has deep causes, such as a lack of money or skill, a diagnostic model like COM-B is the better first step.
?5 questions
Questions people ask
Is the EAST Framework free to use?
Can the EAST Framework be used outside government?
How long do EAST interventions keep working?
Is using the EAST Framework manipulative?
Do I need a behavioural scientist to use EAST?
§12 sources
Sources for the EAST Framework
Behavioural Insights Team (2024). EAST: Four simple ways to apply behavioural insights, revised and updated edition.
Behavioural Insights Team (2024). EAST Framework publication page and history.
Dolan, P., Hallsworth, M., Halpern, D., King, D. and Vlaev, I. (2010). MINDSPACE: Influencing behaviour through public policy. Institute for Government.
NBER Digest (2014). Summary of Hallsworth, List, Metcalfe and Vlaev, The Behavioralist as Tax Collector, NBER Working Paper 20007.
Show all 12 sourcesShow fewer sources
DellaVigna, S. and Linos, E. (2020). RCTs to Scale: Comprehensive Evidence from Two Nudge Units. NBER Working Paper 27594; published in Econometrica, 2022.
Mertens, S., Herberz, M., Hahnel, U. J. J. and Brosch, T. (2022). The effectiveness of nudging: A meta-analysis of choice architecture interventions across behavioral domains. PNAS 119(1).
Maier, M., Bartoš, F., Stanley, T. D., Shanks, D. R., Harris, A. J. L. and Wagenmakers, E.-J. (2022). No evidence for nudging after adjusting for publication bias. PNAS 119(31).
Milkman, K. L. et al. (2022). A 680,000-person megastudy of nudges to encourage vaccination in pharmacies. PNAS 119(6).
Gneezy, U. and Rustichini, A. (2000). A Fine Is a Price. Journal of Legal Studies 29(1).
Madrian, B. C. and Shea, D. F. (2000). The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior. NBER Working Paper 7682.
Michie, S., van Stralen, M. M. and West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science 6:42.
PBS NewsHour / Associated Press (2025). Amazon to pay $2.5 billion to settle FTC allegations it duped customers into enrolling in Prime.
