Start with the current process.
Map the question, the people involved, the data they use and the action they take. Record the baseline before changing the process. A faster dashboard is useful only if its numbers and decisions are dependable.
DIGITAL MATURITY / MY IMPLEMENTATION APPROACH
Find the constraint.
Build from there.
I would not start by buying another platform. I would start with the decision the business needs to make, the evidence available and what is preventing the team from acting on it.
This is my practical marketing-maturity framework, informed by published measurement and maturity research. It is not a certified assessment, an industry percentile or a promise of revenue growth. Public-facing creative alone cannot establish a company’s internal maturity.
01 / DEFINE THE OUTCOME
What should become
easier to decide?
I would agree one priority with the team first: allocating budget, improving lead quality, understanding customer value or deciding whether a tested change should be rolled out.
Separate visibility from readiness.
A brand can have polished creative and weak measurement. It can also have good technical data that no one uses. I would look for evidence across both the technical and organisational foundations.
02 / REVIEW FIVE FOUNDATIONS
Score the evidence.
Keep the unknowns visible.
Select an evidenced stage for each dimension and note what you reviewed. The tool identifies the earliest constraint; it does not average away missing evidence.
1 Captured2 Defined3 Connected4 Tested & used? Not assessed
INDICATIVE READINESS REVIEW
Notes are your evidence references, not independent verification. This tool does not save or submit entries. Unknown means unassessed, not a score of zero. The lowest evidenced stage guides the next investigation; it is not a validated overall maturity level.
03 / TURN THE REVIEW INTO A DELIVERY PLAN
One owner.
One next step.
Clear acceptance criteria.
An illustrative 90-day sequence. I would adjust the timing to the business, data access and team capacity, and move forward when the evidence is ready.
DAYS 1–30 / ESTABLISH THE BASELINE
Agree what the numbers mean.
Work: select the decision, document metric grain and definitions, inspect source coverage, identify owners and log gaps.
Owners: business sponsor, analyst and source-system owner.
Evidence: KPI dictionary, data inventory, baseline reconciliation and a prioritised issue list.
Gate: the team agrees the outcome and every material reporting gap is resolved or explicitly qualified. For an experiment, possible acceptance criteria include representing all eligible assignments and resolving conflicting order IDs.
DAYS 31–60 / CONNECT A SMALL WORKING SLICE
Make one decision repeatable.
Work: connect only the required sources, implement reconciliation checks, define a shared review and make ownership visible.
Owners: engineering for extraction, analytics for validation, marketing for action.
Evidence: a working reporting view, exception log and decision record.
Gate: reconcile the view to agreed sources across a full reporting cycle and explain all material differences. Do not automate an unresolved definition.
DAYS 61–90 / TEST, LEARN AND ADOPT
Check whether the change helps.
Work: pre-register a focused test, agree on the sample and guardrails, measure outcomes and record the decision.
Owners: experiment owner, analyst and business sponsor.
Evidence: valid readout, uncertainties, action taken and follow-up owner.
Gate: validity checks pass, the commercial decision rule is applied and the team can explain why it acted. Scale only after this, then revisit the next constraint.
How I would measure progress
| Measure | Baseline and follow-up method | Evidence to keep |
|---|---|---|
| Data completeness | Represented eligible records ÷ expected eligible records; split by source and, where relevant, experiment arm. Unknown outcomes stay visible. | Counts, exclusions and reconciliation sign-off. |
| Reporting effort | Time comparable cycles before and after; record task scope and error rate. Compare median time, not one convenient run. | Time log and corrections required. |
| Decision adoption | Completed agreed actions ÷ actions due, with a stable definition. A meeting or a dashboard view is not an implemented decision. | Decision register with owner, due date and outcome. |
| Business value | Use a suitable controlled test where feasible; report the effect, interval, cost and margin. Otherwise label before/after changes as descriptive and discuss confounding. | Analysis plan, valid readout and assumptions. |
The stage is a diagnostic. The outcome is the reason to improve.
I would not call the programme successful because its maturity score increased. I would look for more reliable decisions, less avoidable rework and business outcomes measured using an appropriate design.
CONNECTING THIS TO THE CASE STUDY
Fix the foundation
before scaling the test.
The experiment case shows why I would prioritise assignment integrity, outcome completeness and ownership before adding a more sophisticated testing platform.
Read the experiment-integrity case →What I would ask in discovery
- Which decision currently creates the most uncertainty?
- Which metric does the team trust, and how was it checked?
- Where do records disappear or definitions change?
- Who can resolve the issue and who owns the action?
- What evidence would justify the next investment?
FRAMEWORK REFERENCES & LIMITS
The five dimensions, stage anchors and readiness rules above are my practical synthesis, not the BCG/Google scoring model. The sources below informed the emphasis on measurement, skills, ownership and testing. No published uplift estimate is transferred to this portfolio example.