By Ritvars Mētra, Founder of ReadinessCompassADKAR is the most widely used change model in corporate practice, and most of the criticism aimed at it is unfair. It is clear, it is teachable in an hour, and it directs attention at individuals rather than at programme artefacts — which is more than can be said for most of what it competes with.
It also has a boundary, and the boundary is not where people usually look for it. Knowing exactly where ADKAR stops describing your problem is more useful than either defending it or dismissing it.
What it says
Prosci’s ADKAR model describes change at the level of the individual, as five sequential outcomes a person has to reach:
| Element | The question it answers | What failure looks like |
|---|---|---|
| Awareness | Do I know why this is happening? | “Nobody told us anything” |
| Desire | Do I want to take part? | Compliance without engagement |
| Knowledge | Do I know how? | Questions that training should have answered |
| Ability | Can I actually do it, in practice? | People who passed the course and cannot do the task |
| Reinforcement | Will it stick? | Reversion to the old way by month three |
The sequence is the useful part. It is genuinely diagnostic: if people cannot do the task, adding communication will not help, because the gap is at Ability and communication acts on Awareness. That single insight prevents a great deal of wasted effort, and it is why the model has lasted.
The Knowledge/Ability distinction in particular is one most programmes get wrong — it is precisely the gap between having been trained and being able to work.
Where it runs out
ADKAR is a model of an individual’s journey. Every one of its five elements is a property of a person. That is its clarity and its limit, because a large share of what actually stops transformations is not a property of any person.
Consider the readiness literature. Bryan Weiner’s theory of organizational readiness for change describes readiness as having two facets: change commitment — whether people are willing — and change efficacy, the shared judgement that they are collectively capable, given the task demands, the resources available and the situation they are in.
ADKAR covers commitment thoroughly and efficacy only at the level of personal skill. It has no element that represents resources, capacity, competing demands, or structural constraint. So a team can score well across all five — aware, willing, knowledgeable, able, reinforced — and still fail, because there are four other rollouts landing the same quarter and nobody has the hours.
Four failure modes sit entirely outside the model:
- Capacity. Willing, able people with no available hours. ADKAR will read this as a Desire problem, because the observable behaviour is that people are not doing it — and the remedy it suggests will be motivational, which is exactly wrong. Saturation is a system property, not an attitude.
- Incentives. If the measurement and reward system pays for the old behaviour, individual reinforcement cannot overcome it. The constraint is structural.
- Process or system misfit. When the new way genuinely does not work for a real case, resistance is information rather than a gap to be closed.
- Sequencing and dependency. A team can be entirely ready and still blocked because another function is not.
None of this makes ADKAR wrong. It makes it incomplete in a specific, predictable direction: it will tend to convert structural problems into individual ones, because individuals are the only thing it can see. That is a serious risk in practice, because the diagnosis it produces points at the people rather than at the conditions.
The evidence question, asked fairly
ADKAR is a practitioner model. It was developed from consulting practice and benchmarking rather than from a programme of published research, and there is no substantial peer-reviewed literature testing it as a whole.
That is worth stating plainly, but it is not a disqualification, because the same is true of essentially every popular change framework. When Steven Appelbaum and colleagues reviewed fifteen years of literature against Kotter’s eight steps in the Journal of Management Development, they found support for most individual steps but no formal studies covering the model as a whole — and concluded that its popularity derives more from its direct, usable format than from any scientific consensus about results.
Read that as the general condition of the field rather than an indictment of one model. These frameworks are organising devices, not tested theories. They earn their place by making a complicated situation discussable, and they cause harm when they are treated as complete descriptions of what is happening.
Using it well
- Diagnose with it, do not plan with it. Its strength is locating which of five things is missing for a specific group. A communications plan structured as five workstreams named after the letters is the model being used as a template, which is the least valuable thing it does.
- Assess at the group level, and expect different answers. Finance may be blocked at Ability while the warehouse is blocked at Desire. A single organisational ADKAR score averages away the only information you had.
- Test the structural hypotheses first. Before concluding Desire is low, check capacity, incentives and whether the process actually works. If any of those is the real constraint, the individual reading will send you in the wrong direction.
- Measure Ability by observation. It is the one element that cannot be assessed by asking — people reliably overestimate it, and watching four people attempt the task unaided settles it in half a day.
- Do not let Reinforcement mean communication. It means the measurement, incentive and management-attention system, which is usually outside the change team’s control and is the reason it is the most-skipped element.
A fair summary
ADKAR is a good diagnostic for individual-level change gaps and a poor model of organisational constraint. Used to ask “which of these five is missing for this group?” it is one of the more useful things in the practitioner toolkit. Used as a complete account of why a transformation is struggling, it will systematically point at people when the answer is capacity, incentives, sequencing or a process that does not work.
The practical test is simple. When ADKAR says a group is ready and the change still is not happening, the model has not failed — it has told you the constraint is somewhere it cannot see. That is the moment to stop asking about the people and start asking about the conditions.
More on evidence-led readiness in the Change Readiness Hub, or read why no single model explains organisational change.