Maturity Assessment
A maturity assessment scores how advanced an organization is in a given discipline by placing it on a defined progression of stages, from beginning to optimized.
Key takeaways
- Write the stage descriptions first; the questions are derived from observable behaviours.
- Adjective stages cluster everyone in the middle; behavioural stages spread respondents out.
- The output is a stage label plus a description of the next rung.
- Stage assignment routes follow-up: foundational nurture below, sales conversation at the top.
- Self-reported maturity skews high on anything the respondent recognises as good practice.
In depth
A maturity assessment starts from a written model, not from questions. Someone defines four or five stages for a capability and describes the observable behaviours at each one: what a Stage 2 team does with its data, what a Stage 4 team does. Questions are then written backwards from those descriptions, each one testing whether a behaviour is present. The respondent's answers place them at the highest stage whose behaviours they consistently meet, which is why the output is a label rather than a percentage.
The discriminating power of the model decides everything downstream. Stages written as adjectives, such as basic, developing and advanced, push everyone to the middle because nobody knows where the boundary is. Stages written as behaviours spread respondents out. More stages give finer resolution but need more questions to distinguish them, and each added stage makes the top harder to reach honestly. There is also a tension between flattery and usefulness: a model that grades generously feels pleasant and produces no reason to act.
Consultancies have used maturity models in workshops for years; the assessment version puts the same instrument in front of anyone who visits the site. Run as a scorecard quiz, the stage becomes the routing key: Stage 1 and 2 respondents receive the foundational material that describes the next rung, while Stage 4 respondents get a conversation, because their remaining gaps are the ones your product actually solves. The stage description also gives a rep an opening line that is not a pitch.
Maturity models assume a single path, and some organisations are legitimately advanced in one area and deliberately absent in another because their business does not need it. A small firm that will never run a data warehouse is not immature for lacking one. The stage is also a snapshot of self-reported behaviour, so respondents overrate themselves on anything that sounds like good practice. And a model built around your own roadmap will always place the top stage exactly where your feature set sits.
Example in practice
How to measure it
The distribution across stages is the first health check. A working model produces something close to a spread, with a thin top stage and a visible bottom. If two-thirds land on one stage, the questions are not separating behaviours and the thresholds need moving. Track the distribution by traffic source too, since paid and organic audiences often sit at different stages entirely.
Then check whether the stage predicts anything. Compare meeting acceptance and deal size across stages: if higher stages do not convert better, the model is measuring sophistication rather than fit, and the routing built on it is misdirecting sales time. Repeat respondents are the other signal, since someone retaking the assessment a year later shows the stage ladder is being used as a plan.
Common mistakes
The first failure is writing five stages that describe increasing amounts of your product. Respondents notice, and the result reads as a sales argument rather than a diagnosis. Build the model from what capable organisations in the field actually do, including practices your product has nothing to do with. A stage ladder that would still be valid if your company disappeared is the one prospects trust.
The second is stopping at the label. A respondent told they are Stage 2 and nothing more has been diagnosed but not helped, and the result page has no reason to be revisited. Every stage needs the three or four specific moves that reach the next one, named concretely enough to be started this quarter. The gap between where they are and where they could be is what makes them reply.
Frequently asked questions
What does a maturity assessment measure?
It measures how advanced an organization is in a specific capability by placing it on a staged progression rather than giving a pass-or-fail grade. Each stage describes typical behaviors, so respondents see exactly where they stand and what comes next.
How does a maturity assessment qualify leads?
The stage a respondent lands on signals both their needs and their buying readiness. You can route early stages to nurture content and advanced stages to sales, segmenting the pipeline in one session.
How many stages should a maturity model have?
Four or five in most cases. Three is too coarse to give anyone a meaningful next step, since the middle stage absorbs nearly everyone. Six or more forces you to invent distinctions respondents cannot recognise in their own organisation. Pick the number where each stage has a genuinely different set of observable behaviours, then stop.
What is the difference between a maturity and a readiness assessment?
Time horizon and scope. A maturity assessment maps long-term capability across a whole discipline and returns a stage on a growth path. A readiness assessment checks whether the prerequisites for one specific initiative are in place right now and returns a go or not-yet verdict. An organisation can be at a low maturity stage and still be ready for a narrow project.
How many questions does a maturity assessment need?
Roughly two to three per dimension you want to report on, which usually lands between ten and fifteen. Fewer than two per dimension and a single unusual answer can move someone a whole stage. Beyond fifteen, completion drops faster than accuracy improves. Cut dimensions rather than questions if the assessment feels long.
Should we show respondents the stage above theirs?
Yes, and in detail. The stage above is the reason the result is worth reading; without it a label is just a grade. Describe what that stage looks like in daily practice, name the two or three changes that get there, and be honest about what they cost. That description is also the most reused copy in the follow-up sequence.
Can we score dimensions separately as well as overall?
Usually you should. A single overall stage hides the pattern that makes the result useful, which is often strong tooling paired with weak process. Report a stage per dimension alongside the headline stage, and let the weakest dimension drive the recommendation. Sales teams use the per-dimension breakdown more than the overall label, because it names a specific problem.
How do we stop respondents from overrating themselves?
Ask about behaviours instead of opinions. Replace whether your data governance is mature with whether a named owner signs off schema changes, because the second has a factual answer. Offer answer options that describe a real situation rather than a rating scale, and include an unflattering option that is easy to admit. Concrete phrasing pulls self-assessment closer to reality.