Org, people & execution
ADKAR Model
A model of how one person changes: five outcomes, Awareness, Desire, Knowledge, Ability and Reinforcement, reached in sequence for a change to stick. Organisational change succeeds person by person, so diagnose each individual's weakest element and treat that barrier first.
Also known as Prosci ADKAR, Awareness-Desire-Knowledge-Ability-Reinforcement. First set out by Jeff Hiatt (Prosci) in 2003; the primary source is cited in full below.
- Format
- Process / loop
- Level
- Team · Business unit
- Best for
- Plan execution
- Decision stage
- Plan · Execute · Review
- Difficulty
- Introductory
- Time to apply
- An initial per-group assessment takes a day or two; using the model properly spans the whole change, including months of reinforcement after go-live.
Plate · The model
The components
Awareness
The individual understands the nature of the change, why it is being made and the risk of not changing. Awareness of the business reasons, not mere notification.
Signals of strength
People can explain why the change is happening in their own words · The 'why' has come from credible senders, sponsors for business reasons, line managers for personal impact · Rumour and speculation are low
Desire
The individual has made a personal choice to support and engage with the change. Desire responds to personal context and consequence, and cannot be mandated.
Signals of strength
People ask 'how' questions rather than relitigating 'whether' · Visible, consistent sponsorship rather than a launch email · Resistance is being heard and worked, not suppressed
Knowledge
The individual knows how to change: the skills, processes and tools required to operate in the future state, delivered when Desire exists to receive it.
Signals of strength
Training is role-specific and timed close to use · People can describe the new process, not just the old one with new labels · Reference material exists where the work happens
Ability
The individual can actually perform the new way of working to the required standard. Knowledge made real through practice, coaching and time.
Signals of strength
Observed performance, not test scores, meets the standard · Practice environments and floor-walking support exist at go-live · Known gaps between knowing and doing are being coached individually
Reinforcement
Mechanisms that sustain the change after go-live: recognition, measurement, feedback and consequence, so the new way does not erode back to the old.
Signals of strength
Adoption is measured and reviewed months after go-live · Recognition and performance management reference the new behaviours · Workarounds and reversion are noticed and addressed quickly
When it earns its keep
- A rollout is technically complete but adoption is patchy, and you need to diagnose where individuals are stuck rather than re-sending the announcement.
- You are planning a change that depends on people working differently, and want the communications, training and reinforcement designed against what each element actually requires.
- Managers ask what they personally should do during a change; ADKAR gives them a coaching structure for individual conversations.
- A past change has quietly reverted and you want to understand why, typically a missing Reinforcement element.
And when it doesn't
- The change problem is structural or political rather than individual: misaligned incentives, contested strategy or resourcing fights will not yield to individual-level treatment.
- You need an organisational change architecture, sequencing sponsorship, coalitions and momentum across a whole enterprise. Use it alongside an organisational model such as Kotter's, not instead of one.
- The 'change' does not require people to behave differently, for instance a back-end system swap invisible to users. ADKAR adds ceremony without value there.
- The decision itself is still open. ADKAR manages the landing of a decision; it is not a tool for deciding whether the change is right.
How to run it
Before starting, gather the inputs the analysis depends on:
- A precise definition of the change: who must do what differently, from when, and what stays the same.
- Identification of the affected groups and roles, since different groups will stall on different elements.
- A way of hearing where individuals actually are, through surveys, manager conversations or adoption data, rather than assuming.
- Named sponsors and people managers prepared to carry the Awareness and Desire work, which cannot be delegated to a project team.
- 1
Define the change at the level of individual behaviour
State what each affected role must do differently. ADKAR is an individual-level model that rides inside organisational programmes; until the change is expressed as behaviours per role, there is nothing to assess anyone against.
- 2
Assess each group against the five elements
For each affected group, and where it matters each person, score Awareness, Desire, Knowledge, Ability and Reinforcement, typically on a 1-5 scale. Hiatt's rule is to find the first element scoring low, the barrier point, because the elements are sequential: training a person who has no desire to change wastes the training.
- 3
Treat the barrier point, not the symptom
Match the intervention to the stalled element. Awareness needs sponsor communication of why, and why now. Desire needs personal consequence, choice and visible sponsorship, and is the element leaders most often skip past. Knowledge needs training; Ability needs practice, time and coaching, and is where the gap between knowing and doing appears. Reinforcement needs recognition, measurement and consequence after go-live.
- 4
Reassess and keep reinforcing
Re-score after interventions and expect the barrier point to move down the list. Hold Reinforcement in place well beyond go-live; the model's claim is that unreinforced change reverts, so the last element is where sustained adoption is won or lost.
Reading the result
An ADKAR profile per affected group or person, five scores and an identified barrier point, plus a targeted action plan matching interventions to stalled elements, revisited as the change progresses.
- Read left to right and stop at the first weak element; that is the barrier point, and effort spent on later elements is wasted until it clears.
- Different groups stall differently: expect front-line Desire problems and management Reinforcement problems in the same programme.
- Low scores late in the sequence after go-live, particularly Reinforcement, predict quiet reversion even where the launch looked successful.
A worked example
A pharmacy chain rolls out a new dispensing system
A 140-branch UK pharmacy chain is replacing its dispensing system, changing labelling, stock control and the accuracy-check workflow for pharmacists and dispensers. IT deployment is straightforward; the risk is 1,800 people changing daily habits under patient-safety pressure. The change lead runs ADKAR assessments by role across pilot regions before the main rollout.
- Awareness
- Strong for pharmacists, who know the old system falls short of new NHS interoperability requirements. Weak among dispensers, many of whom heard about the change through rota gossip. Action: area managers brief every branch face to face on why, and why now, with the patient-safety case leading.
- Desire
- The barrier point for pharmacists. The new accuracy-check workflow feels slower, and locums see no personal upside. Scores of 2 out of 5 in the pilot. Action: superintendent pharmacist fronts the change, pilot data showing fewer near-miss dispensing errors is shared, and locum onboarding packs and rates are adjusted so locums are not penalised for slower first weeks.
- Knowledge
- Adequate on paper, e-learning completion at 92 percent, but completion is not competence. Action: e-learning trimmed to role-specific paths, and each branch nominates a trained super-user before its go-live weekend.
- Ability
- The pilot's hard lesson: error rates and queue times spiked in week one because staff had knowledge but no practised fluency. Action: a practice sandbox with dummy prescriptions in the fortnight before each branch converts, plus floor-walkers for the first three days and a temporary 20 percent reduction in booked services during conversion week.
- Reinforcement
- The old system remains technically accessible for legacy lookups, and two pilot branches drifted back to running parallel paper checks. Action: legacy access becomes read-only after 30 days, dispensing KPIs move to the new system's reports so there is no alternative record, and area managers recognise branches sustaining clean adoption at 90 days.
The read. The assessment changed the plan: the programme had budgeted heavily for training, the Knowledge element, but the real barriers were pharmacist Desire and post-go-live Reinforcement. Money moved from extra e-learning to sponsorship, locum terms and the 90-day reinforcement routine. The honest caveat is that ADKAR located the barriers; it did not by itself resolve the workload concern behind the Desire scores, which needed a genuine service-reduction decision the model could only point at.
Pitfalls
- Treating communication as Awareness and training as sufficient. Most stalled changes are stuck at Desire or Reinforcement, the two elements programmes fund least.
- Assessing the organisation as one blob. The model's value is in per-group, per-person barrier points; an average ADKAR score hides exactly what you need to see.
- Working the elements in parallel to save time. Training people who lack desire, or reinforcing behaviour people cannot yet perform, wastes the intervention.
- Declaring victory at go-live. Reinforcement is a post-launch activity measured in months, and it is where reversion happens.
- Using ADKAR as the whole change strategy. It is an individual-level model and needs an organisational frame, sponsorship, sequencing and structural alignment, around it.
What the critics say
ADKAR is a proprietary practitioner model whose evidence base rests largely on Prosci's own benchmarking surveys rather than independent peer-reviewed study, a weakness consistent with the broader finding that change management approaches are mostly lacking empirical evidence and rest on unchallenged hypotheses.
By, R. T. (2005) 'Organisational change management: a critical review', Journal of Change Management, 5(4), pp. 369-380.
The claimed change-failure rates used to sell structured change models, including the ubiquitous 70 percent figure, lack valid empirical support, which undermines part of the case made for adopting proprietary methodologies wholesale.
Hughes, M. (2011) 'Do 70 per cent of all organizational change initiatives really fail?', Journal of Change Management, 11(4), pp. 451-464.
The strict linear sequence is an assumption, not an established finding. Real adoption is often iterative and recursive, desire grows from ability and early wins as often as it precedes knowledge, and the model's individual focus can leave structural, cultural and political barriers untreated.
Sources and further reading
- Hiatt, J. M. (2006) ADKAR: A Model for Change in Business, Government and our Community. Loveland, CO: Prosci Research. ↗
- Hiatt, J. M. and Creasey, T. J. (2012) Change Management: The People Side of Change. 2nd edn. Loveland, CO: Prosci Research.
- Prosci, 'The Prosci ADKAR Model'. ↗
- By, R. T. (2005) 'Organisational change management: a critical review', Journal of Change Management, 5(4), pp. 369-380.