ADKAR Model

Theory & origin

Jeff Hiatt published ADKAR in 2003 (formalized in the 2006 book) after studying change patterns across hundreds of organizations at Prosci. His reframe: change programs fail at the individual level, so the unit of analysis should be one person's journey, not the org chart's. The five elements are sequential and cumulative — Awareness of why the change is needed, Desire to participate, Knowledge of how to change, Ability to perform the new way, and Reinforcement to make it stick. The sequence is the diagnostic engine: training (Knowledge) given to someone without Desire is wasted; incentives (Reinforcement) aimed at someone who lacks Ability punish them for a capability gap. The model pairs naturally with top-down frameworks: Kotter moves the organization, ADKAR checks whether any actual humans are moving with it.

Key components

The parts at a glance. Click any term for the full definition, a field example, and the common failure, in the model below.

How a consultant runs it

1. 01 Segment the affected population by role, not org unit: a system change is a different journey for an underwriter than for a branch manager, and each segment gets its own ADKAR profile.
2. 02 Survey or interview each segment against the five elements (1–5 per element) and find the barrier point: the first element scoring low. Everything after it is irrelevant until that one moves.
3. 03 Match the intervention to the barrier: Awareness needs sponsor communication of WHY; Desire needs personal consequence and choice; Knowledge needs training; Ability needs practice, coaching, and time; Reinforcement needs measurement, recognition, and consequence.
4. 04 Assign the right messenger: Prosci's research is unambiguous that business-reason messages land from senior sponsors while personal-impact messages land from direct supervisors. The same message from the wrong mouth fails.
5. 05 Re-profile at each rollout wave: barrier points migrate (Awareness problems in wave one become Ability problems by wave three), so last quarter's intervention plan is already stale.

Worked example

A multifinance lender rolls out a new origination system. Go-live succeeds technically, but sixty days in, half the branches still run shadow spreadsheets. An ADKAR profile by segment finds: underwriters stuck at Ability (trained once, no practice window, reverting under volume pressure), branch managers stuck at Desire (the new dashboard exposes their pipeline to head office for the first time), and tellers fine through Knowledge but with zero Reinforcement (the old screens still work). Three barrier points, three different interventions: a two-week supported-practice window with super-users for underwriters, a sponsor-led conversation about what the visibility is and is not for with branch managers, and decommissioning the legacy screens for tellers. The generic "more training for everyone" plan the PMO had drafted is cancelled — it addressed the one element none of the groups was stuck on.

Common pitfalls

1. 01 Running ADKAR on "the organization" as one blob: the model only works per person or per homogeneous segment
2. 02 Skipping the sequence: interventions aimed past the barrier point (training the unwilling, incentivizing the unable) actively backfire
3. 03 Using the sponsor as messenger for everything: personal-impact messages belong to direct supervisors
4. 04 Stopping measurement at go-live: Reinforcement is an element of the model, not an afterthought, and it is where reversion happens

Sample deliverable

One real engagement, start to finish. Watch the numbers travel from raw input, onto the chart, into the finished artifact.