Why can positive change feel like a loss?

A threatened loss does not need to involve salary, job title or the elimination of a task. Employees have often built value into the way things currently work. The following matrix shows potential associations that a reasonable-looking reform could generate. These are hypotheses to investigate, not universal emotional responses caused by specific changes.

The matrix matters because 'resistance' is an unusually broad label. Someone worried about losing professional autonomy needs a different response from someone who has not learned the new tool. The first situation may require genuine decision-making space. The second may call for practice, support or a better interface. Neither necessarily improves when a campaign says, 'Embrace the change.'

The effects can also run in different directions. A redesign may reduce work for one employee and transfer extra responsibilities to another. An average department satisfaction score can hide both experiences. Segmenting research by role, experience, location, decision rights and everyday situations is therefore often more useful than producing a single engagement figure.

A substantial body of decision research finds that the current option often receives disproportionate weight. In 1988, Samuelson and Zeckhauser described a tendency to retain the status quo even when alternatives are available [1]. One possible reason is that switching creates visible uncertainty and an immediate need to act, while the shortcomings of the current arrangement are familiar and manageable.

Kahneman, Knetsch and Thaler examined the connections between loss aversion, the endowment effect and status quo bias [2]. In certain choices, giving something up carries more weight than gaining something comparable. This does not establish a universal rule that every loss feels exactly twice as significant as every gain. Effect sizes and even the direction of a pattern depend on the reference point, task, context and measurement.

For organisational change, the reference point is especially important. Management compares a proposed process with a desired future: fewer mistakes, lower cost and better forecasting. An employee may compare the same proposal with yesterday: I knew how to do my job, who could help me and when I could use my own judgement. The first comparison highlights potential benefits. The second highlights a possible loss. Both groups can be examining the same proposal in good faith and still reach different conclusions.

People also differ in their preference for routines and responses to imposed change. Oreg developed a measure of individual resistance to change that includes dimensions such as routine seeking and emotional reactions to imposed changes [3]. This helps frame research, but it is not a licence to label critical colleagues with a personality diagnosis. The same person may support a reform when they can influence its design and reject it when the implementation is confusing or unfair.

The point is not to 'bypass' loss aversion. It is to find out which losses people anticipate, whether they are likely to occur and which design choices could reduce them. Better communications and a better change are not interchangeable.

Association research has limits. Imagine that a new digital procedure adds twenty minutes of administration to every shift. 'Extra work' may be a factual outcome rather than an irrational association. If centralisation removes an employee's authority to resolve exceptions while retaining their responsibility for results, reduced autonomy is not just a feeling. If a reorganisation changes the substance of a position, worries about job security may be entirely reasonable.

Such concerns should be investigated directly. Compare tasks, process steps, hours and decision rights. Observe where new friction emerges. Include critical employees in identifying the problem instead of treating them as obstacles to overcome. Ford, Ford and D'Amelio argue that change leaders can contribute to resistance through their own actions and omissions, and that resistance may offer useful information [7]. Their paper develops a theoretical argument, not proof that every objection is correct, but its practical warning matters.

Another distinction is between not wanting and not being able. An employee may fully support the objective but lack access, time, skills or support to perform the new behaviour. A motivational campaign then targets the wrong problem. Conversely, someone may be technically prepared but fear that mistakes will become publicly visible. Further instructions alone will not resolve that concern.

Identity can also matter. Petriglieri conceptualises identity threat as the experience of possible harm to the value, meaning or enactment of an identity [5]. This can help explain why an experienced specialist may not evaluate a new tool neutrally: it may raise questions about what being an expert will mean in the future. It is a theoretical perspective, not a reason to interpret every preference for familiar software as an identity threat.

Many change plans refer to 'the employee' as if there were just one perspective. In reality, the first weeks of implementation may be fundamentally different for employees with different responsibilities. Managers might mostly see the end result, while frontline users experience the transition every day.

These are not measured psychological profiles for each role. They show how to specify research questions. An experienced employee might be the strongest advocate for the new tool; a director may be more worried than anyone on the team. Role is a clue to potential interests, not a diagnosis of a person.

The commercial and operational implication is that one presentation containing five benefits may be inadequate. For each relevant group, identify which situation improves, which existing value needs protection and how the new process can demonstrate that protection. Only then does it make sense to choose a message or intervention.

Imagine a medium-sized logistics business replacing a manually maintained rota with a digital scheduling system. Its leaders expect fewer double-booked shifts, clearer handovers and better visibility of availability. The rota becomes centrally accessible. Employees can submit preferences, and managers can spot staffing gaps earlier. The objective makes sense, and some employees would genuinely benefit.

Yet adoption remains limited. Project managers hear that the system is 'too complicated'. Their first instinct is to make another instruction video. But in this fictional research scenario, that answer may cover several different questions. An experienced planner wonders whether knowledge of unusual exceptions will still count. An employee working irregular shifts wants to know whether personal preferences will actually influence decisions. A team leader worries that a central planner will hold them responsible for outcomes they no longer control. A new colleague primarily wants clarity without having to call someone each time.

'Central system' might therefore be associated with fairness, oversight, distance, monitoring, predictability or dependence. Those meanings could coexist with genuine usability problems. The intervention might be blocked by technical friction and associative resistance at the same time.

The correct approach is not to invent a statistic and claim that 70% of the team fear losing autonomy. No such data exist in this example. Instead, research would begin with interviews about specific moments, free associations with old and new procedures, perhaps a carefully designed implicit association task and observation of actual scheduling behaviour. Employees would then complete realistic tasks: swap a shift, request an exception, verify the rota and find help after an error.

Imagine comparing two variants. Variant A only explains the system more clearly. Variant B also improves exception handling, grants team leaders explicit decision rights and uses experienced planners' expertise to redesign the workflow. Comparing the results could reveal which combination of meaning, process and authority is associated with improved use and reduced friction. A strong causal conclusion would still require an appropriate study design.

A useful questionnaire does more than ask, 'Do you support the change?' If someone ticks 'no', you still do not know whether they anticipate additional work, fear losing influence or encounter an objectively unworkable process. General support questions provide limited explanatory insight without concrete context.

Start with the situation in which behaviour must change. Exactly what must someone do differently, from which date and under what conditions? What was the previous routine? What choices are now required? Use open interviews to explore what employees look forward to, what they fear and what they believe must be preserved. Ask for specific examples rather than general opinions. 'I lose visibility' becomes a researchable statement when you know which handover or screen makes that visibility disappear.

Association research adds a further layer. Ask people for spontaneous words linked to 'new rota', 'old way of working', 'deciding for myself' or 'central scheduling'. Explore which associations are shared within meaningful groups and which differ. Where the research question and design warrant it, an implicit task may help compare the strength of predefined associations. Such a task does not disclose hidden truth about one person's mind. Its validity depends on the stimuli, task, sample and interpretation.

The crucial next step is to compare those associations with actual conditions. Someone who associates the system with 'monitoring' may be reacting appropriately to a new tracking feature. Someone who associates it with 'learning' may still fail because training is inadequate. That is why access, task completion, error correction, support requirements, voluntary use and experienced autonomy should be considered separately. No project needs every measure simply because the measures exist.

Neurofactor's approach is to start with the behavioural question and choose the method combination afterwards. Its research methods overview explains how association research, profiling and behavioural observation can complement one another. EEG or eye tracking is appropriate only if a specific additional research question calls for it. Neither is a compulsory method for understanding resistance to change.

1. Define the target behaviour. Replace 'increase buy-in' with a specific action, such as 'employees submit their availability independently by the agreed deadline'. Identify the people, setting, timing and legitimate exceptions.

2. Establish the reference point. Document how the task currently works, the time it requires and which informal arrangements matter. Without a baseline, deterioration for one group can go unnoticed.

3. Build a gain-and-loss hypothesis for each relevant group. Collect stated concerns, anticipated benefits and associations with both old and new approaches. Do not allow managers alone to determine which themes are investigated.

4. Check whether disadvantages are real. Examine decision rights, workload, schedule predictability, accessibility, safety and how costs and benefits are distributed. A genuine deterioration calls for a substantive solution rather than reworded messaging.

5. Design distinguishable interventions. Compare clearer explanations with process improvements, greater decision latitude or practical support. Avoid changing everything simultaneously if you need to understand what contributed to the outcome.

6. Measure use and experience. Compare behavioural indicators with changes in relevant associations and, where possible, a suitable comparison group or control data. Report meaningful differences between roles and situations rather than relying only on averages.

7. Recheck after the launch period. Early use may reflect novelty or compulsory compliance. Investigate whether behaviour lasts once the programme is routine, how much support remains necessary and whether the promised benefits materialise.

Randomised A/B testing is not always feasible inside an organisation. A phased rollout, repeated before-and-after measurements or comparisons between similar teams may be more practical. The design determines how confidently the results support a causal interpretation. Employee research also requires data minimisation, clear explanations of measurement purposes and appropriate protection of personal information.

Once research clarifies which existing value people fear losing, the response need not be another attempt to persuade. There are several practical routes, depending on the type of loss, whether it is real and the degree of influence employees can exercise.

1. Preserve what genuinely matters. Specify which parts of the current process will remain, such as exception expertise, human support or professional judgement. Do not preserve every habit by default; investigate the function it serves.

2. Offer real influence rather than symbolic consultation. Include relevant employees in decisions where their experience can shape the design. Be explicit about what is open and what is fixed because of safety, legislation or budget. Pretend participation may undermine trust.

3. Replace ambiguity with verifiable information. Instead of declaring 'your role will not change', document the new authority, escalation routes and evaluation criteria. Acknowledge what has not yet been resolved.

4. Protect competence during the transition. Provide safe practice, learning time and help with exceptions. Measure the effort needed to learn the behaviour, not just whether someone watched a training video.

5. Recognise existing expertise. Involve experienced colleagues in scenarios and knowledge transfer. Recognition must influence roles and decisions, rather than consist only of complimentary words.

6. Reduce the size of the first step where responsible. A pilot with genuinely reversible decisions can reduce uncertainty. Explain which outcomes will become permanent and what can be changed after feedback. Never promise reversibility that does not exist.

7. Make the benefit immediate and concrete. Show which recurring inconvenience disappears for each role. 'Digital transformation' is abstract; 'you no longer need to report a late shift swap in three places' is testable.

8. Make the process fair. Examine who receives additional work, who gives up influence and who obtains the benefits. A positive overall business case is not a substitute for a fair allocation of burdens.

9. Measure again in real situations. Check whether negative associations diminish and whether the intended actions become easier, more reliable and sustainable. If attitudes improve but behaviour does not, an important barrier may remain.

Self-determination theory highlights autonomy, competence and relatedness as important for motivation and effective functioning [6]. It does not imply that all organisational changes must be optional or that autonomy alone guarantees success. It does warn against assuming that removing people's control will work simply because the organisational goal is positive.

A better sentence can make a real change easier to understand. It cannot repair a broken process. The following messages refer to the fictional scheduling project and illustrate different communication hypotheses worth testing.

The organisational objective alone: 'Starting Monday, we will use one central scheduling system to improve efficiency and transparency.' The project team understands this, but an employee still does not know what will happen when they need to swap shifts or request an exception.

The employee's action: 'You can enter your availability in one place and see when it has been processed. Help will be available for exceptions during the first few weeks.' This makes the outcome concrete, provided those functions and support are genuinely in place.

Protection of autonomy: 'You can still submit your preferences. The rules for swaps and exceptions appear with every shift, and the team manager retains defined authority when something is unclear.' This addresses a possible loss of influence but is only credible if the process really grants that authority.

Recognition of expertise: 'We have worked with experienced planners to incorporate the exceptions they currently resolve. When the system has no appropriate route, a person will still make the decision.' This can increase trust if the expertise genuinely shaped the system. Otherwise, the promise becomes a further source of scepticism.

Choose among messages by investigating comprehension, associated meaning and behaviour among representative users. Then check which information enables someone to complete the task independently and confidently. If a person feels better after reading the announcement but still cannot submit their availability, the intervention has not solved the practical problem.

A positive change can feel like a loss because people evaluate not only the promise of a future benefit but also the value of what they currently possess. A routine may provide certainty, a role may confer status and an informal agreement may preserve autonomy. Ignoring those existing values means understanding only part of the reaction.

The better question is not, 'How do we convince people that our change is good?' It is: 'What improves for whom, what might actually be lost and what should we therefore redesign?' Study associations, verify the real consequences and measure behaviour. Change becomes credible when its benefits exist not only in a business case but in the practical reality of those expected to adopt it.

Related articles and methods

Neurofactor explains how association research, profiling and behavioural observation can be combined on its research methods page. The applications overview connects those methods with specific organisational questions. Why people believe something is sensible but still do not do it · Communication, process or environment: which behaviour change intervention should you choose?