Nudge theory: real examples, the EAST framework, and where it stops working

Nudge theory arrives in most organisations as a slide about supermarket layouts, and leaves without anything changing. The published examples are duller and considerably more useful.

This article explains what a nudge is, gives real examples with their figures, sets out the framework UK practitioners actually use, and is honest about where the evidence has not held up.

Most workplace nudges are really process design: who does what, by when. See how New Dynamics workflows make those steps explicit.

What a nudge is

A change to how a choice is presented that makes one option more likely, without removing any option or materially changing what it costs. Moving the salad to eye level is a nudge. Banning chips is not. Charging more for chips is not either.

Nudge theory comes out of behavioural economics. What matters for anyone applying it is not the theory but whether a specific change works in a specific setting, which is why the useful sources are the ones that publish their results.

The EAST framework

The Behavioural Insights Team, which began inside UK government, publishes a short framework under an open licence. Its EAST guide puts it in a sentence: "if you want to encourage a behaviour, make it easy, attractive, social and timely (EAST)."

It describes these as "four principles, drawn from behavioural science research and real-world applications", offering "a practical framework for creating effective interventions", while acknowledging that "they don't capture all the complexities of human behaviour".

Examples, with their numbers

Easy. "Harness the power of defaults. People tend to stick with pre-set options, so choose defaults carefully." The guide's example: "presenting renewable energy options as the default increased uptake from 3% to 80% among households in Switzerland, an effect that persisted years later."

Also under easy: "reduce the effort required. Even small barriers can have outsized effects on behaviour. Every extra step or moment of effort required may reduce the likelihood that people will act." The example is the least glamorous and the most instructive, "simply completing university application forms on behalf of US students increased attendance rates from 28% to 36%."

And: "simplify messages... Break a complex goal down into simpler, easier actions." Simplifying court summons forms in New York City "reduced failure-to-appear rates by 13%, preventing thousands of arrest warrants annually."

Attractive. "We are more likely to do something that our attention is drawn towards. Information that is novel and personalised is more likely to be salient." Adding gold coin stickers to a Chinese factory floor "reduced the amount of waste left on the floor by more than 20%". On incentives, savers in the Philippines "increased their savings by 82% when they voluntarily restricted their right to withdraw money until a specific month".

Social. "Show that most people perform the desired behaviour." With an important caution: these approaches "can backfire if the people doing the behaviour are seen as belonging to a different group, or if someone learns that fewer people than they expected are performing the behaviour". Telling people that most colleagues have not completed the training is an anti-nudge.

Timely. Acting when people are receptive, and at the moment the behaviour happens rather than in advance of it.

The honest part

Behavioural science has had a difficult decade, and the practitioners say so. BIT's own preface notes that the field "has attracted more Nobel Prizes – but has also faced major challenges like the replication crisis", and that in revising the guide "we have noted where studies that we mentioned in the first edition have not held up to scrutiny, since this kind of updating is integral to the scientific process".

It also says the guide "will not turn you into an expert on behavioural science".

Three practical consequences:

1. Treat any single published effect as a hypothesis for your setting, not a result you can expect to reproduce.
2. Prefer the boring mechanisms. Defaults and removing steps are more robust than clever framing.
3. Measure what you change, because you cannot tell from the outside whether it worked.

Applying it to workplace processes

Most HR processes fail at the same point: somebody has to do something slightly effortful, at a moment when they are busy.

• Defaults. Meetings default to 30 minutes rather than 60. Holiday requests default to the full entitlement being shown. Pension enrolment already works this way by law.
• Remove steps. Pre-fill what you already know. The strongest example in the guide is literally completing the form for people.
• Simplify the message. One action per email, named, with the deadline in the first line.
• Timing. Send the reminder when the action can be taken, not on Friday afternoon.
• Social, carefully. "Most managers have completed theirs" only works if it is true and if the comparison group is one people identify with.

Our employee productivity post covers the wider point that removing obstacles beats applying pressure.

Where nudging is the wrong tool

• When the obstacle is workload. No reminder fixes a week with no spare hours. HSE's Demands standard expects "adequate and achievable demands in relation to the agreed hours of work"; if that is not met, start there.
• When the policy is the problem. If people avoid a process because it produces bad outcomes, making it easier to use produces more bad outcomes.
• When the decision should be deliberate. Some choices (pension contribution levels, flexible working requests) deserve friction and information rather than a smooth default.
• When you would not tell people. Which brings us to the line.

The ethics line

A nudge is defensible when you would be comfortable explaining it to the people affected. Changing a default and saying so is design. Concealing the alternative, or making the option you prefer the only visible one, is not a nudge, it is a removal of choice with better manners.

The test is simple and unforgiving: publish what you changed and why. If that would embarrass you, do not do it.

Frequently asked questions

What is nudge theory?

An approach to changing behaviour by altering how choices are presented, without removing options or significantly changing costs. It comes from behavioural economics, and practitioners commonly apply it through simple frameworks rather than the underlying theory.

What are examples of nudges?

Published examples include defaulting households to renewable energy, which raised uptake from 3% to 80% in Switzerland; completing application forms on students' behalf, which raised US attendance from 28% to 36%; and simplifying court summons forms in New York City, which cut failure-to-appear rates by 13%.

What is the EAST framework?

Four principles published by the Behavioural Insights Team: make a behaviour easy, attractive, social and timely. It is designed as a practical guide for practitioners rather than a complete account of human behaviour.

Does nudge theory actually work?

Some interventions replicate and some do not. BIT's own guide notes the field has faced the replication crisis and flags earlier studies that have not held up. Defaults and effort reduction are the more robust mechanisms; treat clever framing as a hypothesis to test.

How can HR use nudges?

Change defaults, pre-fill forms, reduce steps, simplify messages to one action each, and time reminders to the moment the action can be taken. These are process design decisions more than psychology.

When should you not use a nudge?

When the real obstacle is workload or a policy that produces bad outcomes, when a decision deserves deliberation rather than a smooth default, and whenever you would be unwilling to tell people what you changed.

Is nudging manipulation?

It depends on disclosure. Changing a default and explaining it is design. Hiding alternatives or engineering a choice you would not admit to is something else, and the practical test is whether you would publish what you did and why.

Your next step: find the step you can delete

1. Pick a process people routinely fail to complete.
2. Count the steps and identify which one you could do for them.
3. Check the default, and whether it is the option you would want chosen.
4. Change one thing, tell people, and measure completion before and after.

For the wider listening and communication side, read and download our employee engagement guide. The guide is free to read, and the PDF uses our short download form.

Want processes where each step has an owner and a due date? Book a New Dynamics demo and bring a process people keep not finishing. You can also email contact@new-dynamics.com.