How do you make desired behaviour appealing without increasing the pressure?

Two interventions can pursue the same goal while suggesting quite different things to participants. The comparison below is a practical design framework rather than a universal hierarchy. Voluntary approaches are not always sufficient, and enforcement is not inherently wrong. Safety-critical activities and legal obligations, for example, require clear standards.

The mistake would be to read the two columns as 'soft' and 'strict'. Clear standards can coexist with autonomy. A supposedly voluntary invitation can still be manipulative if the consequences are unclear. A reward can provide useful feedback, but it may also imply that the task itself has little value.

Looking only at the number of participants cannot reveal these differences. A department may submit forms once because a senior leader requested them. That tells us little about whether useful knowledge will be shared again without reminders. Conversely, slower initial adoption can coexist with genuine commitment among employees who see value in the activity. That is why organisations need measures of quality, repetition and circumstances, not participation numbers alone.

The practical distinction is not between persuasion and silence. It is between increasing pressure on an unchanged activity and making the activity itself more worthwhile, comprehensible and feasible. That usually requires design decisions beyond the wording of an email.

Self-determination theory, developed by Ryan and Deci, highlights three basic psychological needs: autonomy, competence and relatedness [1]. Autonomy concerns experiencing one's behaviour as volitional or aligned with one's own values. Competence concerns feeling capable. Relatedness concerns meaningful connection with others. The theory offers a research-based account of motivation, not a way to forecast every individual decision with certainty.

In organisational communication, autonomy is often reduced to 'give people a choice'. But choosing between two unappealing options is not necessarily meaningful. Nor does autonomy require an organisation to hide the reason for a non-negotiable rule. A manager may rightly insist that customer information is recorded correctly while inviting employees to help design a process that actually works in their day-to-day jobs.

Patall, Cooper and Robinson reviewed 41 studies examining choice and motivational outcomes [2]. The meta-analysis reported generally favourable effects, but also meaningful differences depending on the form of choice and the study conditions. The takeaway is not that every intervention needs another dropdown menu. People need relevant options and a context in which choosing feels possible and worthwhile.

Psychological reactance offers another useful perspective. When a message is experienced as threatening behavioural freedom, some people may respond with anger or counterarguments. Rains reviewed research on the relationship between these responses and reactance [3]. This does not establish that every firm instruction creates resistance. It does suggest that a slogan such as 'those who do not participate are holding us all back' could shift attention away from the value of the behaviour and towards the legitimacy of the demand.

A better design question is therefore not 'How much freedom can we give people before they stop complying?' It is 'Which aspects must remain fixed, and where can participants genuinely decide how to contribute?' Clear professional standards and meaningful discretion are compatible when their rationale is honest and the options are real.

For voluntary participation, that distinction can be decisive. Employees should not necessarily have to share a case at a Friday afternoon meeting if a concise contribution on Tuesday morning is equally useful. The desired outcome is not a particular performance of obedience; it is knowledge reaching the colleagues who need it.

Association research changes the questions you ask. An organisation says 'we want to learn together'. Employees may connect that intention with improvement, trust and professional quality. Yet the very same initiative may also evoke overtime, public exposure, bureaucracy or fear of being blamed. Positive and negative associations can coexist within the same person.

A satisfaction survey alone may not distinguish these responses. Someone can strongly support being a learning organisation while avoiding a weekly contribution. The relevant question is not just whether they approve of learning. It is what they associate with the precise action requested: writing down an error, sharing an unfinished solution, explaining what went wrong or responding in front of colleagues.

That level of specificity matters. 'Knowledge sharing' is an ideal. 'Describe in two minutes what you learnt from this difficult customer conversation' is an observable action in a particular social context. It may involve time pressure and reputational concerns. Research at the moment of action. Ask which words or images come to mind, under what circumstances people would take part and what they think they might lose or gain. Where appropriate, combine open-ended techniques with a structured association task using a validated design.

Do not treat a negative association as a proven cause. If employees associate the system with 'extra work', they may be accurately describing a lengthy workflow. The association may also be common to most new systems, with little ability to explain who actually participates. You need to investigate how common the response is, which conditions evoke it and whether changing the design affects behaviour.

Profiles make the issue clearer. A senior expert may fear losing status, a new colleague may question the quality of their own input and a manager may worry about workload. A generic message explaining 'the benefits for everyone' conceals these distinctions. A useful association map displays attraction and resistance side by side, connected to the relevant action and decision context.

The practical consequences can be surprisingly concrete. If 'learning from colleagues' is appealing but 'being assessed in public' is threatening, a smaller private exchange with clear rules may be a better first step. If the strongest barrier is wasted time, reducing the time requirement matters more than polishing the message. If people think their examples are not good enough, show what a small but useful contribution looks like.

A good intervention should make the promised meaning credible in the actual experience. A message about collaboration cannot compensate indefinitely for a process designed around individual rankings. The association, the environment and the next action need to tell the same story.

Organisations often ask for behaviour because of an outcome that appears much later. More sharing should eventually create a learning culture. Less waste supports sustainability goals. A new workflow will eventually improve efficiency. Those goals can be compelling, but they are not necessarily experienced when a busy employee is asked to take one more step between appointments.

Woolley and Fishbach studied the relationship between immediate rewards and persistence in long-term goals. Across their reported studies, immediate benefits such as enjoyment predicted persistence more strongly than delayed benefits under the examined conditions [4]. That makes immediate value an important consideration in intervention design. It does not show that immediate rewards are sufficient or that everyone responds in the same way.

For knowledge sharing, immediate value may be modest but meaningful. Someone receives a useful response, discovers that a colleague has already solved a similar problem or sees that a tip is incorporated into the team process. The 'reward' is not a voucher. It is direct evidence that the contribution helped. The action starts to matter to the employee, not just to an abstract organisational goal.

A points system sends a different signal. Points may be motivating when experienced as fair, informative feedback. They may also feel like surveillance or competition, particularly when quality matters more than volume. A meta-analysis by Deci, Koestner and Ryan found that certain forms of expected tangible rewards reduced free-choice intrinsic motivation in the experiments reviewed, whereas positive feedback could be beneficial [5]. That is not a blanket rejection of salaries or incentives. It is a warning against assuming every reward is perceived in the same way.

Immediate value must also be real. An organisation promising 'you will see the impact instantly' creates a credibility problem if contributions disappear into a report for six weeks. If employees are asked to share knowledge but never receive a response, the experience undermines the association 'my knowledge makes a difference'.

A useful design rule follows: make the first action worthwhile in its own right, rather than relying only on a distant organisational benefit. For some people that means less effort. For others it means a connection with peers, a sense of progress or the relief of not solving the same problem alone. Which benefit matters should be established with the target group, not guessed from a generic motivational playbook.

Imagine a midsized service company we will call 'North Service Desk'. This organisation and everything that happens in the example are fictional. Employees regularly discuss difficult customer questions, but useful solutions remain in private conversations. Management wants these insights in a shared system. A monthly target is introduced: every employee must submit at least four contributions.

The rationale is understandable. Management expects employees to associate sharing with cooperation. Yet interviews might reveal other possible meanings. An experienced employee could think 'I will have to defend a solution that is not perfect'. A new colleague might think 'I am not qualified to teach anyone yet'. A team leader might hear 'another task that takes time away from customer work'. These are illustrative hypotheses, not actual research findings.

The first design focuses on completeness and visibility. Everyone receives the same guide and contributions appear on a public dashboard. The key management question becomes 'Who has hit the target?' That prioritises volume and visibility before quality. The intervention might increase compliance while strengthening associations with monitoring.

A second design starts with the behaviour itself. After a customer issue is resolved, employees receive a small prompt: 'Could this help a colleague next week? Share the solution in two sentences.' They can also describe the example verbally in a team discussion and have someone else document it. Useful responses from colleagues are visible, but nobody is publicly ranked by name. Everyone knows what happens to the information.

This is more than a softer tone. The action takes less time, appears at a relevant moment, produces a potential immediate benefit and allows meaningful discretion over the form of contribution. The behavioural goal is unchanged: useful information should become available to colleagues. The redesigned process simply connects that goal more closely with professional pride and helping others.

An honest evaluation would measure more than the number of uploads. How many submissions are useful? Are colleagues actually reusing them? Do employees contribute again several weeks later? Are they more or less willing to share mistakes? The two designs could be compared in similar teams, with differences in workload, team size and cross-team influence taken into account. Without evidence, neither design can be labelled superior.

The example illustrates why associational appeal cannot be confined to a campaign slogan. The value promised in the message must be visible in the actual process, social setting and feedback participants experience. Otherwise an organisation may be selling autonomy while managing by surveillance.

A line of copy can be clear to the sender and mean something else to the recipient. Compare several possible invitations for the same fictional organisation. The predicted associations are design hypotheses, not guaranteed reactions.

The warmest-sounding message is not necessarily the most effective. A very loose invitation may protect a sense of freedom while leaving people uncertain about what to do next. A specific request may inspire confidence when the task is genuinely manageable. The objective is not to write the least demanding sentence. It is to combine clear behaviour, recognisable value and appropriate autonomy.

Treat the comparison as a set of hypotheses. Use qualitative research to learn what language evokes resistance and why. In an association study, compare relevant action descriptions, audience segments and contexts. Then test whether changing the message or the workflow changes actual behaviour. If both are altered simultaneously, do not attribute any observed result to copy alone.

This distinction is useful for both behavioural strategists and communication teams. Sometimes the best copy is not a cleverer phrase but a truthful description of a redesigned process. If you promise the task takes two minutes, make sure it actually takes two minutes before testing the headline.

1. Specify one observable action. 'Build a culture of ownership' is not something a researcher can observe directly. 'Share one useful solution after a relevant customer conversation' is. Define who does what, when and in which context. You will need the same precision when assessing results.

2. Find the meaning behind reluctance. Do not stop at 'people are unwilling'. Ask what the requested action seems to say about their role, ability or position. An invitation can suggest trust, monitoring, more work or loss of status. Investigate which meanings matter for which groups.

3. Make the reason relevant without inventing personal motives. Connect the action to values participants genuinely recognise: doing good work, helping colleagues, solving problems more quickly or learning professionally. Also explain why the organisation needs the behaviour. A persuasive story should not conceal the actual purpose.

4. Offer meaningful discretion within clear limits. Let people decide when or how to contribute where that is workable. Keep quality standards, privacy rules and responsibilities explicit. Autonomy is not a vaguely optional programme with no shared expectations.

5. Reduce real friction before polishing the message. If the action is too time-consuming, difficult or poorly integrated into daily work, simplify it. Better copy cannot always compensate for missing time, skills or access. The COM-B model emphasises capability, opportunity and motivation as relevant components of behaviour [6]. This also matters when designing behaviour-change applications.

6. Make immediate value visible. Provide timely, useful feedback, show that contributions are reused or create a small but genuine improvement. Do not choose an arbitrary incentive simply because it is easy to administer. Research what people actually find rewarding.

7. Measure what happens after the launch. Define first participation, voluntary repetition, quality, effects on work and potential harms in advance. Where feasible, compare with a suitable control or a phased introduction. Find out whether positive associations are accompanied by changed behaviour.

Order matters. Teams that have not diagnosed the barrier can rush towards gamification, a new tone of voice or an ambassador campaign. These are tools, not diagnoses. Find out where the action is being blocked, then select an intervention that addresses that obstacle.

A sound study connects three layers: what people say, what the specific action is associated with and what people subsequently do. Each layer adds information. None automatically substitutes for the other two.

Start with an operational definition of behaviour. In a knowledge-sharing project, determine what counts as a useful contribution, who is eligible and over what period it will be assessed. Clicking 'submit' is not the same as sharing a solution that a colleague subsequently uses. Specify success before launching the intervention to avoid redefining it around whichever metric looks best.

Map the associations attached to the behaviour. Interviews, spontaneous association exercises and, when justified, carefully designed response-time tasks can help. Explore links with terms such as 'helpful', 'competent', 'safe', 'waste of time', 'monitoring' and 'proud'. An implicit task is not a truth detector. Results depend on stimuli, contrasts, reliability and the exact outcome being measured. Distinguish between a relatively accessible association and a consciously reported evaluation.

Compare relevant groups and circumstances. Experienced employees, recent hires and team leads may expect different outcomes. Segment by the actual working context and research-based motives, not demographics alone. Test whether your redesign attracts attention, changes the meaning of the action or makes the action easier to complete.

Then choose an appropriate design for evaluating change. Comparable teams could test the original and revised workflows side by side. Measure baseline activity, consider differences in workload and leadership and follow the results for long enough to identify repetition. Individual randomisation may allow people to exchange information across conditions; randomisation by team reduces the number of independent units. These practical details affect the strength of the conclusions.

Watch for unintended effects. Does a leaderboard increase the number of entries while reducing willingness to discuss mistakes? Does a more attractive message boost sign-ups without improving the quality of contributions? Does participation rise but staff workload increase? Those are part of the outcome, not inconvenient footnotes.

The intention-behaviour gap is well documented: plans and action can diverge [7]. That makes 'I would be more likely to participate' a potentially useful intermediate measure, not proof of behaviour change. Trials informed by self-determination theory have shown effects in certain health-behaviour settings, but those results vary and do not automatically transfer to a workplace platform or public campaign [8].

A disciplined report might conclude that the revised process was experienced as more autonomous and that comparable teams generated more useful, repeat contributions. It should not immediately conclude that autonomy alone caused the result. If the redesign also reduced time and improved feedback, those are alternative explanations. Distinguish what you observed from what may explain it and from the decision you now recommend.

You do not need a huge campaign to start learning. A small pilot can help determine whether an intervention deserves a larger test. But there is a crucial difference between 'we ran a pilot' and 'we demonstrated an effect'.

At fictional North Service Desk, two teams could be selected because they handle similar customer questions. One uses the existing monthly request. The other receives a brief invitation after a relevant customer interaction, can choose text or verbal contribution and receives visible, useful feedback. Before either starts, establish normal contribution levels and explore what employees currently associate with knowledge sharing.

During the pilot, record not just submissions but usefulness, time required and whether other colleagues apply the information. Explore the associations at the beginning and end. More voluntary, repeated, useful contributions without higher workload would be an encouraging signal, but differences between teams could still account for some of it.

A follow-up test could isolate mechanisms more carefully. Keep the feedback the same while varying only the level of choice. That would make it easier to distinguish the possible influence of autonomy from timing, convenience and recognition. Define the desired sample size, duration, privacy safeguards and conditions for stopping or adjusting the intervention before starting.

Not participating may also be a legitimate outcome. Some employees have other priorities, and not every behaviour an organisation wants is necessarily in the interests of employees or customers. A responsible study takes the interests of the people affected into account and does not treat every refusal as an obstacle to be engineered away.

The purpose of a pilot is to make the next decision better: scale, revise or stop. Its purpose is not to produce a celebratory chart while the underlying work remains difficult.

When participation stalls, increasing pressure can seem like decisive leadership. Yet pressure does not necessarily make an action more meaningful, safer or easier. In some contexts it makes the existing barrier stronger: monitoring, public judgement, extra work or loss of control.

A better question is not how to explain once more what people should do. It is what value the action can offer them, what currently obscures that value and what the first step can deliver immediately. The final test is practical: do people act, and do they keep acting in the real environment?

Desired behaviour becomes more appealing when meaning, autonomy, competence and practical opportunity reinforce one another. The answer is not simply to apply more pressure, but to design an action that makes sense to the people expected to perform it.