Communication, process or environment: which behaviour change intervention should you choose?

'Encourage more circular behaviour' is too broad to test. 'Visitors return their reusable cup to an authorised collection point before leaving the venue' is much more useful. Even that needs clarification. Which type of cup, which visitors, during which time period, and what happens if a return point is closed or inaccessible?

Map the action as a sequence. A person finishes a drink, decides what to do with the cup, identifies a return location, moves towards it, may wait, and completes the return. Every transition represents a possible failure. Counting cups collected at the end of the evening tells you something about the result, but not necessarily where visitors encounter friction.

Decide in advance what constitutes an observed behavioural outcome. One possible measure is the number of correctly returned cups divided by cups reliably issued into circulation, subject to inventory adjustments for staff use, damaged items, returns from other sessions and reconciliation errors. This is a proposal, not a universally clean KPI. If the denominator is unreliable, an apparently improved return rate may be misleading.

Neurofactor's knowledge base article on the behavioural diagnostic framework helps frame the distinction between a stated intention, an observable action and a business outcome. Those concepts may be connected, but they are not interchangeable. A project needs a clear behavioural target before its intervention options can be judged.

COM-B describes behaviour through capability, opportunity and motivation. These are useful categories, not three independent boxes into which every individual must fit. An inaccessible environment changes what someone expects and perhaps how motivated they feel. Past experiences can influence both confidence and routine.

For capability, ask whether visitors understand the procedure and can perform it. For opportunity, examine station locations, opening times, crowd movements, queues and support. For motivation, explore expectations, habitual responses and meanings associated with returning the cup. Cane, O'Connor and Michie (2012) refined the Theoretical Domains Framework, which can guide more detailed questions about factors shaping behaviour.

The main risk is treating a model as more conclusive than the data. If visitors describe the process as 'difficult', we still do not know whether that means unclear instructions, physical inaccessibility, a faulty payment system or a negative prior experience. Follow up with interviews and observations before making an intervention decision.

Communication, process and environment are practical design categories used in this article. They are not the full list of Behaviour Change Wheel intervention functions and do not represent mutually exclusive scientific mechanisms. They help teams decide who needs to do what next, while the underlying diagnosis remains more detailed.

The matrix below is a working aid, not a validated scoring instrument and not a report of actual results from an event. Each intervention route is a proposition to test. The last column prompts researchers to rule out alternatives before committing to a solution.

One visitor can experience several barriers at once. Another might not feel reluctant at all but simply be unable to find an open return station. The matrix should not be used to classify people into fixed psychological types. Its purpose is to make intervention assumptions visible and checkable.

A practical research team chooses a manageable set of competing explanations, tests the more consequential constraints first and documents the evidence behind each decision. Launching many changes simultaneously may feel decisive, but can make it harder to learn what worked and why.

Communication can be appropriate when people do not understand the procedure, are unsure what is expected or hold an inaccurate belief about the consequences. A short explanation at cup collection might clarify where to return it, how the deposit works and what counts as an authorised point. The message should be accessible near the relevant decision, not only on a screen viewed an hour earlier.

How information is framed can influence comprehension and interpretation. Yet there is no universally superior slogan, colour or tone. Test whether the message is true, understandable and timely. A campaign about being environmentally responsible cannot create a station that is not there. It might strengthen positive meaning but leave the determining constraint untouched.

Social norm claims need particular care. Schultz and colleagues (2007) found in a household energy experiment that descriptive norms can sometimes have unintended effects. Never invent a statement such as 'everyone returns their cup'. Do not publicly shame visitors for circumstances that could be caused by disability, a confusing route or organisational error.

Communication is not inherently a weaker intervention. If the process functions properly and the problem genuinely concerns information, it may be precisely the right first move. But judge it by understanding and subsequent behaviour, not by whether the campaign receives favourable feedback in a presentation.

The process includes the sequence of actions, rules, responsibilities and feedback required for a successful return. It begins with issuing the cup and can continue through collection, confirmation, refund and inventory handling. A cup-return process might create a second queue, require different procedures for different cup types or send visitors between several employees.

Simplification does not mean removing every step. Some steps protect hygiene, payment accuracy, safety or accessibility. The purpose is to identify which actions are necessary and which accidentally create friction. Study what a real visitor must do, rather than assuming that the process diagram in a meeting room represents the actual experience.

If deposits are paid back too slowly, inspect how the system works before designing a promise that refunds are easy. What confirmation appears? What happens when a scanner fails? Can a staff member correct an error, and does the visitor know where to get help? These operational details can strongly shape associations with the entire scheme.

Damschroder and colleagues (2009) describe why organisational conditions matter when putting interventions into practice. A return process that demands extra staff during peak exit times cannot be judged feasible until scheduling and resources have been addressed. What seems frictionless from the visitor side can merely transfer the burden to employees.

The physical environment contributes to what people notice and what is easy to do. Walking routes, lighting, positioning, sound, crowd density and the location of ordinary rubbish bins can shape the behaviour without visitors consciously evaluating each option. If the familiar bin is at the exit and cup returns are hidden around a corner, the default path may be doing much of the work.

Hollands and colleagues (2013, 2017) developed typologies of changes to micro-environments and choice architecture. Their work helps researchers describe interventions that alter proximity, placement and related conditions. It does not establish the effect size of moving a return station in this fictional venue.

A possible redesign places return stations along routes already used by visitors, makes the correct opening recognisable and provides enough space for a safe queue. Any proposed location should also be tested with visitors who use mobility aids, cannot rely solely on visual signage, or travel with children and bags. A solution that improves the average experience while excluding others is not automatically successful.

Mertens and colleagues (2022) reviewed choice-architecture interventions across domains. The research supports taking environmental influences seriously while reminding us that effects differ by technique and context. The legitimate goal is to make a suitable action accessible and clear, not to trick people by hiding reasonable alternatives.

Consider a fictional European event venue with reusable cups. Early in the evening, visitors broadly consider the system reasonable. At closing, staff find abandoned cups on tables. This is an illustrative scenario, not Neurofactor research. No visitor counts, association scores or proven intervention effects exist for this case.

The research team formulates three competing explanations. Hypothesis A is informational: visitors do not know whether a deposit is returned or where to return the cup. A timely, comprehensible instruction may help. Hypothesis B concerns process friction: people anticipate a queue and a slow refund, so shorter and more reliable steps may matter most. Hypothesis C concerns the environment: visitors might be willing to return cups but do not pass a collection point on their ordinary way out.

The team then explores meanings linked to returning a cup. Responsibility and fairness could be attractive associations. Waiting, uncertainty and the possibility of getting the procedure wrong could be negative ones. These are hypotheses to investigate, not findings. The team should neither invent questionnaire results nor treat an illustrative diagram as a recorded dataset.

The decision is now specific. Should the venue explain the procedure, simplify it, relocate stations or combine measures? Interviews, walking-route observation, operational records and a review of deposit handling are needed before one answer is justified.

An association map is not a creative brief by itself. If 'hassle' appears prominent, that does not tell designers to use a greener poster, a cheerful face or a larger arrow. First identify which part of the action may evoke that meaning, which competing explanations exist and how a proposed change would plausibly influence it.

Write a functional design principle: 'Visitors can return cups on a normal exit route without unnecessary uncertainty about the destination or deposit.' From this principle follow several prototypes: repositioned stations, an improved refund sequence, a clear completion signal, or combinations. The principle is stable while the form remains open to testing.

State the evidence needed for each hypothesis. Information problems call for comprehension checks and behavioural observations. Placement problems require route mapping, visibility measures and actual returns. Process friction requires completion rates, delays, errors and staff workload. Do not choose a measure simply because it is easy for the analytics dashboard to display.

The knowledge base discusses triangulation in research. Different methods answer different questions and can contradict one another. That disagreement can be valuable: visitors may say they understood a sign while their walking route reveals they never saw it.

A good evaluation plan is not the one with the most numbers. It is one that distinguishes meaningful behavioural change from measurement errors and unwanted consequences. The following matrix is a planning example for the fictional case, not a table of measured results.

Where feasible, compare suitable event sessions or venue areas using a pre-specified outcome and a credible allocation method. Randomisation may be possible by session or entrance area, but visitors and staff can move between conditions. A simple before-and-after comparison may be distorted by programme differences, crowd size or weather.

Moore and colleagues (2015) explain why evaluating complex interventions requires attention to implementation, mechanisms and context. Glasgow and colleagues' RE-AIM framework (1999) adds questions about reach, effectiveness, adoption, implementation and maintenance. The later G-069 article addresses real-world intervention testing in depth. Here the principle is simpler: build the evaluation into the intervention choice, rather than adding it after launch.

If observations show that stations are nearly invisible, a placement change might be the best first experiment. If people misunderstand an otherwise working system, communication may be more useful. If both problems exist, a combined prototype could be justified. The choice depends on what the evidence supports, what is feasible and how much can be learned.

More components do not always mean better performance. Squires and colleagues (2014) reviewed whether multifaceted healthcare interventions outperform single-component approaches; the findings do not justify a universal 'more is better' rule. Adding several measures may make a project harder to implement and conceal which change generated an observed effect.

Use a decision record that names the suspected barrier, evidence, competing explanations, proposed mechanism, intervention components, costs, risks, primary outcome and stopping rule. Include uncertainty plainly. A polished workshop presentation should not make a provisional hypothesis look like a settled conclusion.

Even a successful pilot may not be sustainable. Volunteers can make a temporary return point appear easy to operate while normal staffing cannot absorb the extra effort. Implementation conditions must be included from the outset, not discovered after scale-up.

A guest rushing for the final train may face a different constraint from someone staying at the venue. A person with reduced mobility may be unable to take an extra route. Friends may return several cups together while the deposit system assumes individual transactions. These circumstances require practical investigation rather than stereotypes about age, personality or commitment.

Segmenting by the actual situation can be more useful than assigning fixed behavioural types. Examine which exits visitors use, when queues form, how instructions are encountered and which accessibility needs arise. The same person may behave differently on a quiet afternoon and after a crowded concert. A situational pattern is not a permanent character trait.

Language and cultural familiarity may affect understanding, but do not assume deficits in a demographic group without evidence. Test instructions and pictograms with the relevant visitors. When a proposed intervention only works for those who can read small signs or walk long distances, improve the design instead of labelling others uncooperative.

Staff are also a target group in implementation, even if the desired behaviour concerns visitors. A change that saves guests ten seconds but overwhelms employees or creates errors may not be worthwhile. Include front-line teams in the design and process evaluation.

The first error is mistaking positive intentions for demonstrated behaviour. Sheeran and Webb (2016) explain why that transition warrants separate investigation. The second is defining behaviour so vaguely that success cannot be measured. The third is interpreting associations as hidden individual causes rather than testable explanations.

The fourth error is using a moral appeal to compensate for an inaccessible process. The fifth is changing several things without documenting their implementation. The sixth is interpreting a higher count of returned cups as success without checking issued volumes, duplicates, stock movement or the context of the event.

The seventh is ignoring unintended harm. A solution might raise the average return rate while making a doorway unsafe, excluding visitors with disabilities or creating a disproportionate staff burden. These concerns belong in the primary decision, not in an optional disclaimer at the end.

Avoiding these errors does not require extravagant claims about the brain. It requires careful measurement, transparent reasoning and a willingness to discard a creative idea when the diagnosis does not support it.

A practical method does not need to be enormous, but it does require explicit assumptions and decisions. The following steps connect behavioural analysis with design and evaluation, while leaving room for a project to revisit earlier conclusions.

1. Define the action. Specify who should do what, where and when, and record relevant exceptions.
2. Observe the journey. Identify breakdowns and check capability, opportunity, timing and actual process defects.
3. Investigate meanings. Examine positive and negative associations and test them against interviews and context.
4. State the proposed mechanism. Explain why communication, process, environment or policy could remove the barrier.
5. Build alternatives. Prototype at least two defensible approaches and plan for accessibility and delivery.
6. Pre-specify outcomes. Define behaviour, secondary effects, cost, a comparison condition and stopping rules.
7. Test and revise carefully. Document what was implemented, interpret uncertainty and only scale when the evidence warrants it.

This turns an association map into a useful input for decisions. It also leaves an audit trail explaining what was changed, why that change was plausible and how its impact will be judged.

Communication, process and environment can all influence behaviour. Their value depends on which constraint is present. When that is unknown, teams often recommend the intervention that happens to match their own skill set. A campaign can obscure a broken process; a new layout can leave distrust unchanged.

Association research helps reveal possible meanings and resistance. Behavioural diagnosis examines whether a desired action is feasible. Design creates alternatives. Evaluation determines what changed, for whom, at what cost and with which possible harms.

The useful question is not which solution will look best in the presentation. It is which change removes the most relevant, evidenced barrier. That is how insight about human choice becomes a responsible programme of behavioural design.