By Ritvars Mētra, Founder of ReadinessCompassThe change maturity assessment came back at 2.4. The improvement plan, unsurprisingly, is to reach 3.0 within eighteen months, and there are now workstreams for each dimension that scored below the target.
Eighteen months later the organisation will very likely be at 3.0, because organisations that set out to reach a maturity level generally reach it. Whether anything about how change actually lands will have improved is a separate question, and the model has no way of answering it.
A descriptive instrument used prescriptively
This is the structural problem, and it is documented in the wider maturity-model literature rather than being a peculiarity of change management.
Evaluating business process maturity models against established design principles, Röglinger and colleagues found that the models handle basic and descriptive design principles adequately — they can characterise where an organisation currently sits — but that the principles for prescriptive use are hardly met. A systematic review of their limitations reaches a related conclusion: research in the field concentrates on producing new models rather than validating existing ones, and empirical evidence of their usefulness in practice is scarce. Maturity models have been criticised as step-by-step recipes that simplify business reality.
In plain terms: these instruments are reasonably good at telling you where you are and were never designed to tell you what to do. Almost every organisation uses them for the second purpose.
What the level actually measures
| What the assessment scores | What you wanted to know |
|---|---|
| A documented change methodology exists | Whether anyone follows it under deadline pressure |
| Impact assessments are performed | Whether their findings changed a single decision |
| Sponsors are identified and briefed | Whether a sponsor has ever held a date |
| Change resources are trained and certified | Whether managers do the new thing themselves |
| Readiness is measured before go-live | Whether the measure could have returned a bad answer |
Every item in the left column is an artefact. Artefacts are countable, auditable and comparable across business units, which is precisely why maturity models use them — and precisely why the score can rise without the organisation getting better at change.
Then the level becomes the target
The moment a maturity level appears in an objective, two well-documented effects begin.
The first is Campbell’s Law, stated by Donald Campbell in 1979: the more any quantitative social indicator is used for decision-making, the more subject it becomes to corruption pressures, and the more apt it is to distort the process it was intended to monitor. The cheapest route from 2.4 to 3.0 is to produce the missing artefacts. That route will be taken, because it is available, measurable and finishes on time.
The second is surrogation — named by Willie Choi, Gary Hecht and William Tayler in The Accounting Review — where people stop treating a measure as a representation of a construct and begin acting as though the measure is the construct. “Being good at change” quietly becomes “being at level 3”. Their research found surrogation is worst when a single measure carries the construct, which is exactly the shape of a maturity score.
The combination is why maturity programmes so often produce a well-documented change function that is no better at landing change. The score was honest when it was first taken. It stopped being honest the moment it became the goal — the same trap that makes a readiness criterion worthless if it can be met without the underlying thing being true.
The deeper mismatch
Maturity models assume capability is general and cumulative: that an organisation is at a level, and that raising the level improves everything.
Change capability does not behave that way. A company can be genuinely excellent at ERP rollouts and hopeless at anything requiring sales behaviour to shift, because the constraints are different: one is mandatory and capability-bound, the other discretionary and incentive-bound. A single organisational level averages those into a number that describes neither.
Nor is the binding constraint usually a missing practice. It is more often capacity, incentives, or a sponsor who will not hold a date — none of which is fixed by adopting another process, and all of which can coexist with a high maturity score.
What to do instead
- Assess against outcomes, not artefacts. Take the last three changes. Did they land? What actually stopped the ones that did not? Three post-mortems tell you more than any assessment instrument, and they name causes rather than gaps.
- Diagnose per change type. Mandatory system changes, discretionary tool adoption and process changes fail for different reasons. One score cannot cover them.
- Fix the binding constraint. If sponsors will not hold dates, no amount of documented method helps. If teams have no capacity, another template is a tax.
- If you must have a score, carry it on several measures that can disagree — artefacts, outcomes and behaviour together. Surrogation is hardest where one number cannot stand in for the whole.
Where they do help
Maturity models are genuinely useful for what they were designed to do. As a structured conversation about what exists and what does not, in an organisation that has never thought about change capability, they surface real gaps quickly. As a common vocabulary across business units, they help. As a baseline taken once and never turned into a target, they are honest.
The failure is not the instrument. It is asking a descriptive tool to serve as an improvement plan, and then managing to the number it produces.
More on frameworks and evidence in the Change Readiness Hub, or read why no single model explains organisational change.