Kotter's change process
Kotter's change process is a sequence of eight stages for leading a transformation across a whole organisation, published by John P. Kotter in Leading Change (1996). Kotter built it by inverting eight errors he had observed in around a hundred attempted transformations: each stage corrects one of those errors. The first three create the climate, namely the sense of urgency, the guiding coalition, the vision and the strategy. The next three engage the organisation through communication, the removal of obstacles and the first wins. The last two consolidate the gains and then anchor the new practices in the culture. The sequence addresses the leader who carries the transformation; the business analyst contributes on the vision, the communication material, the process obstacles and the measurement of the wins.
Goal
Kotter's change process leads a transformation through eight stages to be cleared in order. The question it answers is one of sequence: in what order to commit the actions so that the change holds once the programme ends. It is a leadership framework with a beginning and an end.
The deliverable is a change management plan structured by the eight stages: for each stage, what the organisation commits, who answers for it and the observable sign that counts it as cleared. Kotter publishes no template; this plan is the form in which what each stage demands becomes verifiable.
The order matters because the stages rest on one another. A vision developed without felt urgency mobilises nobody; short-term wins obtained without a shared vision read as isolated improvements. Kotter describes skipping a stage as the main failure mechanism: it gives the impression of speeding things up and takes away the support the change was resting on.
Force Field Analysis weighs the forces driving a change against those restraining it at a single point in time; Kotter's process works over the life of the change and orders the passage of the whole organisation through it.
Usage
When to use it
- Organisation-wide transformation: a merger, a new service model, the replacement of a core system touching several business lines.
- Buy-in as the binding constraint: the solution is settled and carried by senior management; whether people adopt it will decide the result.
- A long programme carried by a leadership mandate: the sequence assumes a coalition that stays in place from the first work on urgency through to anchoring.
- Diagnosis of a stalled change: re-reading the eight stages names the one that was skipped or handled too fast, which indicates what to do next without establishing the cause of the failure.
- Framing the change management strand of a project: it supplies the milestones and the exit criteria expected before go-live.
When not to use it
- Local or purely technical change: an eight-stage programme costs more than it returns; prefer a Force Field Analysis and a one-page adoption plan.
- Continuous transformation with no end point: the sequence closes once the culture is anchored; the restatement Kotter published in 2012 addresses this case and describes a permanent capability.
Description
Eight stages, three phases
The eight stages group into three phases, a grouping Kotter's firm introduced afterwards to present the model. Three stages run before any visible result appears, which asks of senior management a patience many programmes do not have.
Creating a climate for change: stages 1 to 3
Stage 1, establishing a sense of urgency. The threshold Kotter sets is high and most attempts miss it from the outset: leaders confuse a little urgency with urgency. Urgency comes from putting in front of employees the data that makes the status quo untenable: market-share figures, a comparison with a competitor, customer complaints read out as they stand, a dated regulatory deadline. The stage is cleared when a large majority of the management line holds keeping the current situation to be riskier than the change. Two traps: a diffuse anxiety paralyses the room without directing it; a manufactured crisis is paid for in credibility as soon as it is seen through.
Stage 2, creating the guiding coalition. Kotter asks the coalition for four qualities held together, which one person rarely has: the position power to decide, expertise in the domain affected, credibility with the people concerned and the capacity to lead. The third obliges bringing in people whom those on the ground listen to, beyond the executive committee. The stage is cleared when the group can settle a hard question in session and its decision is received without being renegotiated in the corridors. The common trap is a steering committee made of delegates without a mandate, which reports and does not decide.
Stage 3, developing a vision and strategy. The vision states the target state in a form simple enough to be laid out in under five minutes and to draw a reaction of understanding. The strategy says how to get there. Visioning is the technique the business analyst brings to this stage. The stage is cleared when every member of the coalition tells the same story with no slide. Without a vision, a programme fragments into a list of unconnected projects, each defensible and the whole unreadable; a detailed plan and a budget then occupy the space without ever giving direction.
Engaging and enabling the organisation: stages 4 to 6
Stage 4, communicating the change vision. Kotter's finding is quantitative: the communication of a transformation is under-delivered by a factor of ten, even a hundred. A vision is repeated on every channel and in every routine exchange, project reviews included. Every visible act of a leader confirms or contradicts the message, and a behaviour that contradicts it cancels months of communication. The stage is cleared when employees far from the programme give the vision back in their own words. A short survey measures that.
Stage 5, empowering employees for broad-based action. This stage consists of removing what stops people who are already convinced from acting: a structure that partitions, a job description that forbids acting as the vision asks, an appraisal system that still rewards the old behaviour, a constraint in the IT system. Process Analysis feeds this stage by showing where the real work collides with the rule. Interviews on resistance with the people concerned complete the picture. The deliverable is an obstacle register, each line with an owner and a date for removal. The trap is the exhortation to autonomy that leaves intact the mechanisms penalising it.
Stage 6, generating short-term wins. A useful win is planned, visible, unambiguous and attributable to the change, obtained within six to eighteen months. Its function is to create the credibility the rest of the programme needs: it proves right those who committed and deprives the sceptics of their argument. Metrics and Key Performance Indicators serve to establish the baseline value before the intervention, without which the win stays unverifiable. Two traps recur: rebranding as a win a result that would have arrived anyway and choosing an improvement nobody outside the programme notices.
Implementing and sustaining the change: stages 7 and 8
Stage 7, consolidating gains and producing more change. This is the stage most often cut short. Declaring victory after the first successes releases the pressure at the moment when the change has touched only the surface, and the old practices return of their own accord. Kotter asks the opposite: use the credibility earned to attack the heavy structures left aside, launch a new wave of projects, bring in reinforcements and renew the composition of the coalition in stages. This phase usually runs longer than the six before it put together.
Stage 8, anchoring new approaches in the culture. The culture moves last, once the results are in and tied to the behaviours that produced them. That tying is done out loud and repeatedly, failing which everyone attributes the result to something else. Anchoring then runs through personnel decisions: the change survives when it determines who is hired, promoted and appointed to key posts. A successor who never bought in undoes in a quarter what the programme took three years to build.
Kotter's later reformulation
Kotter revisited his model in 2012 in the article Accelerate!, expanded in 2014 into the book of the same name. There he presents the eight items as accelerators that continuously feed a second structure, network-like and built on volunteers, running alongside the hierarchy. The vocabulary changes with the logic: enlist a volunteer army, enable action by removing barriers, sustain acceleration. This version describes a continuous capability where the 1996 version describes a sequential programme with an end. Both formulations circulate under the same eight-step heading. The 1996 version has a settled, citable text; the 2012 one describes an arrangement its author has gone on reworking since. Mixing the two vocabularies in one plan blurs the logic of each.
The model's limits
The process presents itself as a fixed sequence, whereas organisational change advances by going back over ground: the stages overlap, several run in parallel and a programme that holds returns regularly to urgency or to the vision. Read as a checklist, the model promises a linearity the field does not deliver.
It is then built from the top down. It assumes a coalition with enough position power to carry the organisation, and it has almost nothing to say about a change driven from the base or by peers. Where that power is not assembled, a Stakeholder Matrix establishes who could carry it. The 2012 restatement answers part of this criticism.
Kotter drew it from observing around a hundred transformations he worked on as a consultant, with no experimental protocol and no comparison group. It is a leadership practitioner's synthesis, widely adopted, to which no measured result is attached. The "proven approach" label, common in its commercial distribution, rests on no published study.
Its structure, finally, is self-confirming. Since every failure is explained by a stage skipped or rushed, no observation can contradict it. The model is a good design tool and a poor evidence tool.
AI considerations
A language model is useful on the stages that produce text in volume. Stage 4 is the type case: the vision has to be adapted for clinical staff, for managers and for external partners, in the languages the organisation works in, then repeated on every channel. That volume is what explains the chronic under-delivery of the stage. The assistant also produces a first list of obstacles for stage 5, from the existing job descriptions, procedures and appraisal criteria. It works through the free-text answers of a perception survey to show where the vision has not landed and prepares plausible objections ahead of a staff assembly.
Judging urgency escapes it: that requires reading the mood of a room and knowing the history of previous changes in this organisation; an urgency message drafted by a machine reads like a press release and produces the opposite effect. Composing the coalition rests on credibility and trust, which no organisation chart yields. The vision, finally, has to be carried: a formulation no leader can repeat without reading it does not clear stage 3, whatever its literary quality.
The free-text answers of a perception survey and the interviews on resistance at stage 5 are personal data within the meaning of the Federal Act on Data Protection, and sending them into an external service exposes remarks made under cover of anonymity, which shuts off the source best placed to report on the programme's real state.
Examples
A cantonal hospital group in French-speaking Switzerland is replacing the paper patient file with an electronic patient record, within the Swiss electronic patient record (EPR) programme. The concept illustrated is the exit criterion.
Change management plan
Move to the electronic patient record
| Stage | What is committed | Observable exit criterion | Deadline |
|---|---|---|---|
| 1. Urgency | Twelve months of transcription incidents presented in every department, with the cantonal deadline for connecting to the EPR. | Heads of department quote their own department's incident count with no slide. | M0 to M2 |
| 2. Coalition | A group of six: medical directorate, director of nursing, IT director, a head of emergency medicine, a ward nursing manager, the liaison from the cantonal health department. | The group settles the freeze of the CHF 180'000 paper budget in session without the decision being reopened. | M1 to M3 |
| 3. Vision | A one-sentence vision: every piece of information useful to the patient available at the patient's bedside, current and readable by the next department. | The six members of the coalition lay it out in under five minutes, with no notes. | M3 to M4 |
| 4. Communication | Vision taken up at the opening of every department meeting, in the staff newsletter and during the medical directorate's visits, in French and in German. | Perception survey: 80% of healthcare professionals give the vision back in their own words. | M4 to M14 |
| 5. Empowerment | Four obstacles removed: duplicate paper entry still required by internal regulations, absence of tablets on the wards, read permissions siloed by department, training for cover staff. | Obstacle register closed, every line dated and signed by its owner. | M5 to M12 |
| 6. Wins | Emergency department live on the electronic record, first phase costed at CHF 450'000. | Duplicate laboratory orders: 9.4% before, 3.1% after, measured over three months. | M6 to M12 |
| 7. Consolidation | Extension to surgery and internal medicine, redesign of the admissions pathway, two new members joining the coalition. | Two further departments live and the admissions pathway redefined, with no announcement that the project has ended. | M12 to M30 |
| 8. Anchoring | Command of the electronic record written into the job descriptions for nursing management posts and into the induction programme for new employees. | Three consecutive management appointments made on the criterion; no return to paper found in internal audit. | M24 to M36 |
The exit criterion column is what separates this plan from a project schedule. Can an observer outside the programme establish the fact without being told what was undertaken to obtain it?
The deadlines of this programme give the measure of the effort: four months pass before the vision exists and twelve before the first measured result. Stage 7 alone runs eighteen months. It is also the window where the programme is most exposed: the moment the duplicates drop from 9.4% to 3.1% is the moment a hurried management would announce the project a success. These durations belong to this programme; Kotter dates only the short-term wins, six to eighteen months.
Visualisations
The sequence unfolds in time. Its position on an axis, its grouping into three phases and the loss of momentum that threatens stage 7 are read in space, which makes it a diagram.
The change management plan is made of values: a stage, a commitment, a criterion, a deadline. The table is the deliverable itself, the one a coalition keeps current and against which it checks whether it has the right to move to the next stage.
Cost
| Phase | Level | Justification |
|---|---|---|
| Preparation | High | Gathering the data that makes the urgency tangible and assembling a coalition with the power to decide and credibility on the ground. This work is counted in weeks and depends on arbitrations the analyst does not make. |
| Execution | High | The sequence occupies the coalition continuously, from the first work on urgency through to anchoring in the culture. The stage 4 communication on its own is a permanent effort. |
| Documentation | Medium | A one-page vision, an eight-stage plan with its exit criteria, an obstacle register and a dashboard of the wins. Little volume, but an update every time a stage is cleared. |
Tooling
The plan itself fits in a spreadsheet or on a page of the organisation's collaboration space: eight rows, the commitment, the exit criterion, the owner and the deadline. The format matters less than its accessibility to the whole coalition and its revision at each stage cleared. A project management tool becomes appropriate as soon as stages 5 and 7 generate enough tasks to deserve tracking. The obstacle register sits there as a work package, with owners and dates.
Stage 4 plays out on the existing channels: intranet, staff newsletter, team meetings, staff assemblies, group messaging. Repetition across all these channels calls for a communication calendar someone keeps. A short anonymous survey tool, run two or three times during the programme, measures whether the vision has landed and signals where resistance concentrates. Stage 6 demands a measurement setup: a dashboard in the existing management tool, fed from the business systems, with the baseline value. Collaborative whiteboards (Miro, Mural, FigJam) serve the vision workshops and the obstacle register, particularly for a coalition spread over several sites.
Sources
- Kotter, J. P. (1995), Leading Change: Why Transformation Efforts Fail, Harvard Business Review, 73(2), 59-67: the original article, where the eight errors observed in around a hundred transformations are stated and inverted into eight stages.
- Kotter, J. P. (1996), Leading Change, Boston, Harvard Business School Press; reissued 2012 by Harvard Business Review Press with a new preface and an unchanged text: the full account of the eight-stage process, the reference source for the stage names and for what each one demands.
- Kotter, J. P. (2012), Accelerate!, Harvard Business Review, November 2012: the restatement of the eight items as accelerators of a second, network-like structure operating alongside the hierarchy.
- Kotter, J. P. (2014), Accelerate: Building Strategic Agility for a Faster-Moving World, Boston, Harvard Business Review Press: the book-length development of the 2012 article, with the vocabulary particular to that version.