Decision capability assessment
Most maturity models average away the constraint. This one finds what could block the marketing decision you need to make. Eight dimensions, ten trade-offs, about four minutes. No email required to see your result.
Free diagnostic
Most maturity models average away the constraint. This one finds what could block your decision.
A team with a pristine data pipeline and a finance director who ignores the output is not ready to reallocate a budget. It is stalled. The useful question is not how sophisticated the measurement stack looks. It is how ready you are to use evidence to make a particular decision, and what is currently preventing you.
This assessment does three things. It asks which decision matters. It places you on eight dimensions we use in client work. Then it puts you through forced trade-offs to work out what you would actually protect under pressure.
- One decision so the same profile can mean ready for annual planning and not for weekly reallocation.
- Eight dimensions from decision clarity through ownership, scored foundational to embedded.
- Ten trade-offs between the things a decision system can optimise for, adapting as you answer.
- A plan naming the constraint on that decision, and the three moves to make in order, starting with defining the decision, not assembling three years of data.
Nothing is sent anywhere until you choose to send it. Your answers stay in this browser tab.
How the trade-off scoring works
We do not ask you to rank your priorities, because asked directly everybody wants measurement that is fast, rigorous, transparent, cheap, comprehensive and owned internally. Direct ranking measures how agreeable the list is, not what you would defend.
Instead you see two priorities at a time and say which you would protect, on a five-point scale from “clearly this” through “leaning” and “genuinely torn”. Those answers fit a graded Bradley-Terry model: each priority is given a latent worth, and the chance of preferring one over another is a logistic function of the difference between them. Strength and ties are encoded as observed probabilities of 1, 0.75, 0.5, 0.25 or 0 (not as a binary win/loss), so “slightly prefer” carries less weight than “clearly prefer”. Because ten comparisons across eight items is not much data, the fit is ridge-penalised, which shrinks the estimates toward each other and stops any priority that happened to win every time it appeared from running away.
The order is adaptive. After each answer the model refits and picks the next pair it can least predict, since a comparison between two items you rate equally carries the most information. Where too few comparisons separate two priorities, the result groups them into a shared tier rather than printing a rank order the data will not support.
This is the same reasoning we apply to measurement generally: state the model, state what it assumes, and be explicit about how much the evidence can carry. If that sounds like the way you would like your marketing measured, that is rather the point.
Want the thinking behind the assessment?
The full field guide covers all of it: maturity, stakeholder mapping, the data audit, build versus buy, scenario planning, validation, and the four ways in-housed measurement quietly dies.