Normalisation of behaviour: meaning and application

A new way of working or service usually starts with attention, but at some point you want people simply to do it. When is that point reached, and how do you know? This article explains where the concept comes from, how it differs from adoption, habit and social norm, how to measure normalisation and what communication does differently then.

Normalisation of behaviour is the process in which new behaviour becomes so much a part of everyday routines and ways of working that it no longer requires separate attention, motivation or recruitment.

Normalization Process Theory (NPT) describes how a practice becomes embedded in an organisation through people's shared work. Habit research describes how repetition in a stable context makes individual behaviour more automatic. Normalisation does not persist by itself: if the context changes, the behaviour can lapse.

This article follows the meaning of normalisation in implementation research. In 2009, Carl May and Tracy Finch set out Normalization Process Theory. Their starting point is the question of how practices become routinely embedded in everyday life, and why this succeeds for some practices but not for others (May & Finch, 2009).

Its central proposition is that practices become routinely embedded, or normalised, because people work individually and collectively to enact them. That investment has to be made continuously (May et al., 2009). NPT distinguishes four kinds of work, described simply here:

1. Coherence: understanding what the new practice is, how it differs from what went before and what it means.
2. Cognitive participation: engaging people and getting them to decide to take part.
3. Collective action: actually carrying out the practice, with the necessary skills, resources and coordination.
4. Reflexive monitoring: appraising how the practice turns out and adjusting it where needed.

A systematic review found 130 papers from 108 studies that used NPT in implementing healthcare interventions. Researchers found the theory useful, with criticism mainly of its terminology. The authors state that their findings apply only to healthcare (May et al., 2018). Applying NPT to communication or HR questions therefore needs its own support.

Habit research describes what happens in the individual. Wood and Neal describe habits as learned associations between a response and features of the context in which it was repeatedly performed, such as a location, a preceding action or particular people. Once formed, the habit can be triggered by that context without a mediating conscious goal (Wood & Neal, 2007). Habits are therefore relatively insensitive to short-term goal changes (Wood & Rünger, 2016).

Lally and colleagues asked 96 volunteers to perform an eating, drinking or activity behaviour daily in the same context for twelve weeks. Time to 95 per cent of each person's maximum automaticity ranged from 18 to 254 days. Missing one opportunity did not materially affect the process (Lally et al., 2010). The often quoted median of 66 days comes from this study, among participants for whom the model fitted well. A recent systematic review found medians of 59 to 66 days in four studies, with a range of 4 to 335 days, based on a limited evidence base (Singh et al., 2024).

Context is crucial here. Life transitions, such as a new job or moving house, reduce people's exposure to the cues that trigger old habits (Wood & Rünger, 2016). Such a moment opens space for new behaviour, but can also loosen a newly formed routine.

Adoption is a core concept in Everett Rogers's diffusion theory, on how innovations spread through communication within a social system (Rogers, 2003). Normalisation is about what happens next: does the behaviour last, and does it become embedded? A habit can be part of this, but normalisation also needs an environment that carries the behaviour: agreements, resources, roles and work processes. A social norm can strengthen normalisation, but behaviour can also be normal without anyone experiencing it as a norm. Perceived social norms and social proof have their own articles.

You usually combine three kinds of information.

Perceived normalisation and the work around it. Finch and colleagues developed the NoMAD instrument for NPT. The final version has 23 items: 20 items on the four mechanisms and three general questions. These ask how familiar something feels when you use it, whether it is currently a normal part of your work and whether it will become one. Validation used 831 questionnaires from six health-related sites; reliability per mechanism ranged from 0.65 to 0.81 (Cronbach's alpha) (Finch et al., 2018). The authors deliberately give no fixed scoring rule or cut-off, allow adapted wording and list test-retest reliability, among others, as still to be studied.

Automaticity in the individual. Verplanken and Orbell's Self-Report Habit Index (SRHI) contains twelve items on repetition, automaticity and identity (Verplanken & Orbell, 2003).

Behaviour over time. Usage data show whether behaviour stays stable without extra prompts. Frequent behaviour is not yet a habit; habit is more than how often someone did something (Verplanken & Orbell, 2003). Interviews and observation show which work is stalling, as in many NPT process evaluations (May et al., 2018).

Early on, communication revolves around attention, explanation and recruitment. The emphases below are an editorial translation of NPT and habit research: testable starting points, not researched campaign rules.

• Recruit less. Persuading people who already take part may suggest the behaviour is still special. Aim recruitment at those not yet taking part.
• Support more (collective action). Keep resources, instructions and agreements in order and resolve bottlenecks quickly.
• Fixed moments and cues (habit). Link the behaviour to a stable context, such as a set time or place.
• Make it visible (reflexive monitoring). Share how the behaviour is going and what has been adjusted.
• Explain again when things change (coherence). Newcomers and changed circumstances call for renewed explanation.

Normalisation is therefore not a fixed final phase of every customer journey or campaign; whether and when behaviour becomes normal depends on context and embedding. See also how to translate a target group profile into a communication strategy.

1. Describe the behaviour concretely. Who does what, when and in which context?
2. Determine the level. Is it about individual habits, embedding in a team or organisation, or both?
3. Take a baseline. Record use, felt familiarity and the four mechanisms.
4. Adapt instruments carefully. Rephrase NoMAD items, for example, and test them first; the validation applied to healthcare settings.
5. Follow over time. Measure at several points and record context changes, such as a new season, a reorganisation or a move.
6. Find the cause before campaigning. If the behaviour stalls, determine per mechanism where the work gets stuck. A behavioural diagnostic framework helps to organise causes.

See also how to turn target group insights into testable messages.

This example is fictional and contains no research results.

Situation. An office organisation with three sites in one city introduces shared bikes for trips between them. After a launch campaign some employees use the bikes, but use drops once reminders stop.

Decision question. Should the organisation launch a new recruitment campaign, or do something else?

Available information. Usage data per site, reports of flat batteries and signs that new employees do not know about the scheme.

Suitable approach. The research team runs a short questionnaire with adapted NoMAD items, such as 'Does using the shared bike feel like a normal part of your work?', and a few automaticity questions from the SRHI. It also talks with employees and the facilities team.

Possible interpretation. If employees understand why the scheme exists (coherence) and want to take part (cognitive participation), but the bikes are often not ready (collective action), the bottleneck lies in support, not motivation. New employees who do not know the scheme need explanation during onboarding.

Next step. The organisation sets measurement points and measures in advance, repeats the measurement after a change of season and checks whether use stays stable without reminders.

• Mistaking adoption for normalisation. A quick start says little about what happens once attention fades.
• Assuming a fixed period. The 66 days are a median from one study with a wide range, not a norm for every behaviour or group.
• Forgetting context. A move, reorganisation or change of season can interrupt normalised behaviour. A lapse is then not automatically unwillingness.
• Confusing individual and organisation. A team can embed a practice while individual members feel little automaticity, and vice versa. An organisational average does not predict one person's behaviour.
• Transferring findings untested. NPT and NoMAD have mainly been studied in healthcare, habit research mainly with health behaviours and self-report. Other contexts need their own testing.
• Seeing normalisation as good by definition. Undesirable behaviour can become normal too, so first judge whether the behaviour is desirable.

Normalisation of behaviour requires continuous work by people and organisations (Normalization Process Theory) and repetition in a stable context (habit research). Keep normalisation separate from adoption, habit and social norm, measure at several levels and over time, and allow for context and lapses. In this phase, communication means recruiting less and supporting and making visible more.