From Resistance to Adoption: Six Conditions for Lasting Transformation

Deploying a technology, announcing a strategy or training employees does not guarantee that working practices will change. Adoption lasts only when six conditions come together: meaning, value, capability, opportunity, trust and reinforcement.

The technology has been deployed. Employees have completed their training. Leaders have presented the vision. The programme dashboard shows healthy numbers: accounts activated, sessions delivered and users connected.

Yet several months later, the old spreadsheets are still circulating. Teams continue to rely on familiar processes. Managers use the new system during demonstrations but not consistently when making day-to-day decisions.

The transformation appears to be progressing, but the work itself has barely changed.

This happens because organisations frequently confuse five different stages:

1. Deployment: the solution becomes available.
2. Use: employees begin experimenting with it.
3. Adoption: the new practice becomes regular and useful.
4. Appropriation: teams integrate and adapt it to their work.
5. Institutionalisation: the behaviour becomes part of how the organisation operates.

Completing the technical project does not complete the transformation. It often marks the point at which the more difficult organisational work begins.

Adoption is not primarily a communication problem

When adoption falls short of expectations, organisations tend to respond in predictable ways. They increase communication, schedule more training or ask managers to "bring their teams on the journey".

These interventions may help, but they will not resolve problems rooted in work design, incentives, power, confidence or trust.

Employees may understand the logic of a transformation perfectly well while also observing that:

• the new tool makes their work more complicated;
• performance objectives still reward the old process;
• their manager does not use the solution;
• there is no time to learn through practice;
• mistakes could damage their credibility;
• the transformation threatens their expertise or autonomy;
• the expected behaviours receive no meaningful recognition.

What leaders call resistance may therefore be a rational response to the signals sent by the organisation.

Recent research reinforces this interpretation. After examining 51 successful enterprise AI deployments, Stanford's Digital Economy Lab concluded that differences in outcomes were not primarily explained by the technology. Organisational readiness, processes, leadership and the willingness to learn and fail were more consequential.

Research from the University of Cambridge reaches a related conclusion. Using panel data from the UK Management and Expectations Survey, researchers found that firms with stronger existing management practices were significantly more likely to adopt AI. Performance measurement capabilities appeared to be particularly important.

Adoption should therefore be treated as a property of the organisational system, not simply as an individual decision.

Six conditions shape that system.

1. Meaning: "I understand why this matters"

The first condition is a credible interpretation of why change is necessary.

Employees need to understand:

• why the current situation is no longer sufficient;
• what problem the organisation is trying to solve;
• what needs to change;
• what should be preserved;
• what the transformation means for their work;
• how their contribution fits into the wider ambition.

Karl Weick's work on sensemaking is essential here. People do not passively receive meaning from leadership presentations. They construct it through experience, observation and conversations with others.

A transformation narrative becomes credible when it connects five elements:

1. a change in the external or internal environment;
2. a strategic problem or opportunity;
3. the organisation's chosen response;
4. the behaviours that must change;
5. the practical implications for each affected group.

A corporate presentation about becoming "AI-first" or "future-ready" does not necessarily answer the employee's questions:

• What should I do differently next week?
• Which decisions will change?
• What responsibility will I retain?
• How will my performance be assessed?
• What happens to the expertise I have spent years developing?

When these questions remain unanswered, employees create their own explanations. The transformation may then be interpreted as a cost-cutting exercise, a surveillance programme or another temporary management initiative.

The CEO question

Can employees explain, in their own words, why the transformation is necessary and what it changes in their work?

Repeating the official message is not evidence of sensemaking. The stronger indicator is whether teams can translate the strategy into their operational reality.

2. Value: "This improves something that matters"

Understanding a transformation does not mean that people will adopt it.

Everett Rogers showed that the diffusion of an innovation depends partly on its perceived relative advantage. Users need to believe that the new approach is better than the one it replaces.

Fred Davis's Technology Acceptance Model similarly highlights perceived usefulness and perceived ease of use.

The difficulty is that value is not experienced uniformly.

For the organisation, a new system may promise:

• higher productivity;
• greater standardisation;
• lower costs;
• better visibility;
• stronger data collection;
• faster decision-making.

For the employee, the same system may mean:

• additional data entry;
• reduced autonomy;
• increased monitoring;
• the loss of valued responsibilities;
• new expectations without a reduced workload;
• the devaluation of existing expertise.

Strategic value for the organisation does not automatically translate into practical value for the user.

Leaders should therefore examine value at several levels:

• strategic value for the enterprise;
• operational value for the team;
• professional value for the individual;
• value for the customer;
• social or symbolic value.

This analysis also needs to recognise potential losses. Adoption decisions are influenced not only by what people expect to gain, but also by what they believe they may lose: status, expertise, control, relationships or professional identity.

The CEO question

What concrete benefit can each affected group associate with the new behaviour?

The answer will rarely be identical for executives, managers, specialists and early-career employees.

3. Capability: "I know how to do this"

Even when employees understand the transformation and recognise its value, they still need the ability to act.

Capability includes more than technical knowledge. It encompasses:

• functional skills;
• process understanding;
• professional judgement;
• confidence in one's own ability;
• access to support;
• opportunities to practise;
• time to learn;
• feedback on real work.

The COM-B model developed by Susan Michie and her colleagues proposes that behaviour depends on capability, opportunity and motivation. An organisation can communicate persuasively and create strong motivation while still failing to produce the expected behaviour if people lack the necessary capability.

Traditional training has limitations in this respect. It often teaches features in an environment separated from actual work. Employees learn what the system can do, but not necessarily how to use it within complex decisions, competing priorities and real organisational constraints.

Capability becomes operational through application, repetition and feedback.

Research from MIT's Humans in the Loop project adds an important warning. Generative AI can enable an employee to complete a task without retaining or developing the knowledge required to perform it independently. This is a form of cognitive offloading.

The relevant question is therefore not only whether AI helps someone complete a task faster. Leaders must also ask whether its use:

• builds capability;
• preserves professional judgement;
• supports progressive autonomy;
• or creates a dependency that remains hidden until the system fails.

The CEO question

Do employees merely know how to operate the tool, or do they understand the work, decisions and risks it is changing?

Lasting adoption should strengthen collective capability rather than produce a temporary increase in output at the expense of future expertise.

4. Opportunity: "The organisation allows me to act"

A person may understand the transformation, value it and possess the required skills, yet still be unable to change their behaviour.

The organisation must provide a genuine opportunity to act.

Opportunity includes:

• time;
• access to information and data;
• suitable tools;
• decision rights;
• workable processes;
• management support;
• coordination across functions;
• regulatory permission;
• realistic workloads;
• compatible objectives.

Organisations routinely ask for behaviours that their systems make difficult.

A company may encourage experimentation while using a budgeting process that penalises deviation. It may promote collaboration while measuring individual performance exclusively. It may celebrate autonomy while requiring decisions to move through several layers of approval.

In these situations, the formal message and the operating system are in conflict.

The Cambridge research on organisational capital and AI adoption is particularly relevant. It suggests that technologies require specific organisational complements. AI does not automatically compensate for weak management systems; it may amplify their limitations.

This is why adoption cannot be separated from organisational design. If the desired behaviour conflicts with workflows, governance or incentives, communication alone will not resolve the contradiction.

The CEO question

If an employee wanted to adopt the new behaviour immediately, which process, rule or dependency would prevent them from doing so?

This question shifts the focus from employee reluctance to organisational responsibility.

5. Trust: "I can try this without putting myself at unacceptable risk"

Adoption requires people to take risks.

Trying a new method may require an employee to:

• admit that they do not yet understand something;
• ask for help;
• produce an imperfect result;
• question an established practice;
• challenge a decision;
• report a problem;
• experiment without certainty of success.

Amy Edmondson defines psychological safety as a shared belief that a team is safe for interpersonal risk-taking. Psychological safety does not mean permanent comfort, low standards or the absence of accountability. It means that employees can ask questions, acknowledge mistakes and raise concerns without expecting humiliation or disproportionate punishment.

In an AI-related transformation, trust has several dimensions:

• trust in the reliability of the technology;
• trust in the quality of the data;
• trust in the governance framework;
• trust in how employee information will be used;
• trust in leadership's intentions;
• trust in the future of one's role;
• trust in the manager's response when something goes wrong.

The World Economic Forum identifies human accountability and transparency-driven trust as important conditions for AI adoption at scale.

Deloitte's 2026 Human Capital Trends research goes further by describing culture as infrastructure. When organisations neglect culture while transforming technology and work, they accumulate "culture debt": unresolved tensions that make future change slower, more fragile and less credible.

The CEO question

What does an employee genuinely risk by experimenting, making a mistake or raising a concern?

The answer is found less in official policy than in how the organisation responded the last time someone did so.

6. Reinforcement: "The organisation rewards the new behaviour"

The final condition concerns what happens after someone begins to adopt the new practice.

Will the behaviour continue if:

• the manager returns to the old method;
• performance indicators remain unchanged;
• promotions reward different priorities;
• only short-term results receive recognition;
• legacy systems remain mandatory;
• senior leaders bypass the rules themselves?

Edgar Schein's work shows how leaders shape culture through what they consistently pay attention to, measure, resource, reward and sanction.

An organisation's real priorities are therefore visible in its allocation of attention and recognition.

Reinforcement can take several forms:

• leadership role-modelling;
• management feedback;
• recognition;
• performance indicators;
• objectives;
• resource allocation;
• recruitment criteria;
• promotion decisions;
• rituals;
• organisational stories.

Repeated reinforcement turns a new behaviour into a routine and eventually into a norm. Without it, teams return to familiar practices that feel safer, simpler or more advantageous.

This is also where leaders frequently undermine their own transformation. They describe one set of behaviours as strategically important while continuing to reward another.

The CEO question

Are people who adopt the new behaviour more likely to succeed in the organisation than those who preserve the old one?

If the answer is no, the cultural system is contradicting the transformation.

The six conditions operate as a system

These conditions should not be treated as six independent workstreams. They reinforce one another.

An average score can be misleading. A transformation may perform well across five conditions and still fail because the sixth is critically weak.

Excellent training cannot compensate for an incompatible process. An inspiring vision cannot compensate for a lack of trust. A valuable technology cannot compensate for incentives that reward the old way of working.

Diagnosing an adoption problem

Before increasing communication or commissioning another round of training, an executive team should ask six questions:

1. Meaning: Do affected groups understand the strategic reason for the change?
2. Value: Can they identify a concrete benefit in their own context?
3. Capability: Do they possess the required skills and judgement?
4. Opportunity: Do processes, resources and decision rights allow them to act?
5. Trust: Can they experiment and report problems without disproportionate risk?
6. Reinforcement: Do leaders, metrics and rewards support the new behaviour?

Each condition can be assessed on a scale from one to five. The purpose, however, is not to produce an attractive aggregate score. It is to identify the system's breaking point.

A weak condition should lead to a specific organisational decision:

• clarify the narrative;
• redesign the value proposition;
• build a capability;
• change a process;
• create safer conditions for experimentation;
• align performance indicators;
• change a leadership behaviour.

The CEO's role

The CEO does not need to manage every adoption activity. The CEO does need to ensure that the organisation is not asking employees to adopt behaviours that its own systems prevent or penalise.

This means:

• connecting transformation to strategy;
• defining observable behaviours;
• resolving contradictions between functions;
• giving managers genuine authority and resources;
• protecting learning and experimentation;
• aligning incentives and recognition;
• monitoring adoption rather than deployment alone.

It also means accepting a less comfortable possibility: weak adoption may reveal a flaw in the transformation itself.

Feedback from employees is not always a barrier to overcome. It may expose poor design, insufficient value, unmanaged risk or consequences that senior leadership failed to anticipate.

Conclusion: Adoption is an organisational outcome

A transformation becomes sustainable when the new behaviour is understood, valuable, achievable, permitted, safe enough to attempt and consistently reinforced.

When change is not adopted, the first response should not be to label employees as resistant or automatically intensify communication.

The first question should be:

Which condition for adoption has our organisation failed to create?

That question changes the nature of the conversation. Adoption is no longer treated as an employee attitude problem. It becomes a leadership and organisational design responsibility.

And that is often where genuine transformation begins.

Selected references

• Chris Argyris and Donald A. Schon, Organizational Learning.
• Fred D. Davis, "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology".
• Amy C. Edmondson, "Psychological Safety and Learning Behavior in Work Teams".
• Susan Michie, Maartje M. van Stralen and Robert West, "The Behaviour Change Wheel".
• Everett M. Rogers, Diffusion of Innovations.
• Edgar H. Schein and Peter A. Schein, Organizational Culture and Leadership.
• Karl E. Weick, Sensemaking in Organizations.
• Bennett School of Public Policy, University of Cambridge, More Than Just Plug and Play, 2026.
• MIT Industrial Performance Center, Humans in the Loop, 2026.
• Stanford Digital Economy Lab, The Enterprise AI Playbook, 2026.
• World Economic Forum, Organizational Transformation in the Age of AI, 2026.