What the science says

Consensus

It helps to separate what the model claims from whether the claims are true, because the first is widely known and the second is where the trouble lies. The model, as consolidated by Prochaska, DiClemente and Norcross (1992) from earlier smoking-cessation work (Prochaska and DiClemente, 1983), pictures change as a journey through five stages: precontemplation, where a person has no intention of changing; contemplation, where they are weighing it up; preparation, where they intend to act soon and may have made a first attempt; action, where they are actively changing the behaviour; and maintenance, where they are working to hold the change in place, with relapse treated as a normal part of the cycle rather than a failure. Around this sit a set of "processes of change" said to matter more at some stages than others, a "decisional balance" of pros against cons, and a person's confidence, or self-efficacy. The model's central practical prescription is stage-matching: work out where someone is and give them the intervention that fits that stage.

‍

The part that is genuinely and durably valuable, and worth stating clearly before the criticism, is the readiness insight. The model insisted that "not ready to change" is not the same as "ready but failing", and that the two need completely different handling. That mattered, because the field it entered tended to design everything for people ready to act, when in reality most people with a habit to change are nowhere near that point; the 1992 paper made the pointed observation that since most smokers are pre-contemplative, programmes offering only action-oriented help will "underserve, misserve, or not serve" the majority. Add to that an intuitive, teachable vocabulary that gives practitioners a shared way to talk about where someone is, and you have the reasons the model spread so far through public health and health promotion, and why it is still taught today.

Controversies

This is where the honesty has to be blunt, because the Transtheoretical Model is one of the most criticised frameworks in the whole of behaviour change, and the criticism goes to its core rather than its edges. The first and most damaging point is that the stages are not real, distinct states at all. They are arbitrary thresholds cut across a continuous underlying variable, roughly, how ready or close to acting someone is, and the cut-points are essentially conventions. Sutton (2001) showed this plainly, and West (2005) put the standard example sharply: a smoker planning to quit within thirty days, who has made a quit attempt in the past year, is filed under "preparation", while the same person planning to quit in thirty-one days is only in "contemplation", which means the boundaries are, in West's phrase, "arbitrary lines in the sand", and the confident statements you often hear, that some percentage of a population is "in the contemplation stage", carry little real meaning. The second point is that people do not move through the stages in the orderly sequence the model implies; they skip, stall and cycle. The third, and the one that matters most for anyone spending money on this, is that the model's central promise fails to deliver: the stage-matching hypothesis, that tailoring an intervention to someone's stage beats not doing so, has weak and inconsistent support. A stage-outcome correlation, West (2005) notes, says no more than the obvious, that people who are already thinking about changing are more likely to try than people who are not. A BMJ systematic review of stage-based interventions for smoking found "only limited evidence" for their effectiveness despite their widespread and uncritical use (Riemsma et al., 2003), and Sutton (2001) concluded the evidence base was "meagre and inconsistent". The overall verdict from the critics is that the model's success rests on face validity, it feels right and it is easy to teach, far more than on the evidence, which is why West (2005) argued it was time to put it "to rest".

Limitations

Beyond the core dispute, three practical limits follow. The model labels and describes far more than it explains; knowing someone is "in contemplation" does not tell you what will actually move them. Its central construct is hard to measure reliably: the various stage questionnaires do not cleanly capture discrete stages, and different methods of assigning people to stages agree poorly with each other (Sutton, 2001). And by concentrating on conscious readiness and decision-making, it neglects the reward, punishment and associative learning that drive so much habitual behaviour, the automatic processes that often operate beneath deliberate weighing of costs and benefits (West, 2005). One further, quieter limitation is that its very teachability lets it persist and be marketed long after the evidence turned, which is a hazard in its own right.

Open questions

The field has largely moved toward treating readiness as a continuum rather than a set of boxes, and toward more dynamic accounts of motivation, of which West's own PRIME theory is one example. The open questions are whether any of the model's specific apparatus, the named stages, the processes, the matching rule, adds real predictive value over a simple question like "on a scale of nought to ten, how ready are you?", and, more sympathetically, whether the readiness heuristic still earns its keep even if the formal stage-model is wrong, by stopping practitioners from making the genuine mistake of pushing action on people who have not begun to consider it. There is also good evidence that tailoring interventions in other ways, to people's triggers and motives, outperforms both untailored approaches and stage-matching (West, 2005), which points to where the effort is better spent.

So what

The practical instruction here is unusually specific: keep two things from this model and let go of the rest. Keep the readiness insight, that "not ready" is a real state needing different handling from "ready", and keep the shared vocabulary as a rough and convenient way to talk about where someone is. Let go of the belief that the five stages are real, distinct states, and, more importantly, let go of the idea that matching an intervention to someone's stage is a proven, high-value lever. Treat readiness as a rough continuum, a rough sense of how close someone is to acting, rather than a set of five boxes with a validated engine behind them.

‍

For companies, this bears on any behaviour-change product or communication, from health and finance apps to sustainability nudges and internal change management. The readiness insight is real and worth designing around: someone who has not decided to change needs a different message, the "why", the information, the motivation, than someone ready to act, who needs the "how", the tools, the plan, the removal of friction. Using that split as a segmentation heuristic is perfectly sound. What is not sound is building an elaborate, rigid five-stage funnel and claiming, or believing, that stage-matching is a proven multiplier, because the evidence does not support the machinery. A single honest readiness question will do more for you than a validated-sounding stage algorithm that is not actually validated.

‍

For political parties and campaigns, the same logic applies to mobilisation. It is a mistake to treat every non-voter or non-supporter as ready but unpersuaded; some have never considered the ask, some are weighing it, some are ready to act, and each needs a different approach. The readiness framing is a genuinely useful way to segment a population and avoid wasting an action-oriented appeal on people who are not there yet. But it remains a heuristic, not a validated pipeline, so it is not worth building a rigid stage-matched operation in the expectation of a matching effect that the research has struggled to find.

‍

For government and public health, this is the model's home ground, and it is where the caution matters most, because a great deal of public money went into stage-matched interventions whose evaluations came back mixed to negative (Riemsma et al., 2003). The honest guidance is to keep the readiness insight, tailoring to whether people are even considering a change is sensible, while recognising that stage-matching as a designed intervention lever is not well supported, so it is not a sound basis for funding elaborate stage-algorithm programmes on the promise of a matching effect. Simpler continuum-based tailoring, or a more diagnostic framework such as capability-opportunity-motivation, may do as well or better. This is one of those cases where the most useful thing a practitioner can be told is what not to over-invest in.

‍

There is a quiet dual-use point here that is really about honesty. A model this intuitive and this easy to teach can carry on being used, and sold, long after its own field has largely concluded that its central mechanism does not work, precisely because familiarity and face validity are persuasive in themselves. The responsible move is to be clear about the line between the useful heuristic and the overclaimed mechanism, and not to trade on the model's familiarity to sell a certainty it has not earned.

‍

The summary is that the Stages of Change gave behaviour-change work one real and lasting gift, the recognition that readiness matters and that "not ready" is its own state deserving its own response, wrapped inside a model whose specific machinery, the real discrete stages, the orderly progression, the payoff from stage-matching, has not survived the evidence, to the point that a leading figure in the field called for retiring it. So use it as the evidence licenses: as a rough, teachable heuristic for meeting people where they are, not as a validated engine of change. Keep the readiness insight, hold the stages lightly, and never mistake a model's intuitive appeal for proof that it works.

‍