Decide what is worth building.Build it,secure it,and make it stick.
Data IQ™ is how every Data Products engagement runs: one method and one senior team from the first question to a system your people use, with the reasoning written down at each step, so nothing has to be reconstructed when the work changes hands.
The senior people who frame the problem stay accountable through delivery.
One engagementFive phasesFour stagesSecure runs through all five
Phase 1 of 5AssessmentDecideOn file: readiness scorecard
Phase 2 of 5StrategyDecideOn file: ranked shortlist
Phase 3 of 5Architecture & buildBuildOn file: Decision Brief, signed
Phase 4 of 5DeploymentBuild, Make it stickOn file: decision log
Phase 5 of 5EnablementMake it stickOn file: runbook and handover
One engagement, phase by phase: the stages at work and what goes on file. Decide spans phases 1 and 2, Build 3 and 4, Make it stick 4 and 5; Secure runs through all five. Illustrative.
1
Four stages say what the work is. Five phases say when it happens.
Phase boundaries are where engagements lose their context: a strategy team leaves, a delivery team arrives, and the reasoning behind every decision has to be reconstructed. Here, one team carries the work across them. The stages are groupings, not steps in a queue: most engagements start in one and pull in two more, and security and governance run through every phase.
Eight practices, one method. The team that writes your roadmap builds the system it describes, which is the argument for hiring one firm instead of four.
2
Two documents decide most of what follows.
One decides what is worth building. The other fixes, before any engineering, what the system has to decide and how it will be judged.
The Data IQ™ decision lens
Three questions, seven tests
Every idea is ranked against the same tests, so the order survives the meeting.
Is it worth it?
01Business value: the number it moves, and what that is worth
07Measurable outcome: whether you could tell, in a number, that it worked
Can we deliver it?
02Data readiness: whether the data exists, and is usable
03Technical feasibility: what it takes to build, and whether it has been done
04Operating fit: who would run it, and whether the process can absorb it
Can we deploy it?
05Risk and governance: what has to be true for legal, risk and security to say yes
06Adoption burden: who has to change how they work, and how much
Six dimensions decide whether AI survives production. All six are published.
This is the AI-readiness view of the Data IQ™ model. The free assessment scores it from your answers; the Assessment Sprint scores the same six against your systems. The complete model, applied in the Sprint and in our engagements, reaches wider.
Every dimension is checked four ways: is it documented, owned, measured and tested? A dimension that scores low points to the one engagement that fixes it, so the result is a decision, not a report.
Dimension
What the Sprint checks
If it scores low
Data foundation
Lineage traced on your actual pipelines; retrieval tested against the questions people really ask.
Scores the six from your answers. Free: eighteen questions, about seven minutes, no email required. Your weakest dimension points to one place to start.
Scores the six against your systems. Two weeks: the six dimensions scored with evidence, a gap register, an order-of-magnitude cost view and prioritized next steps. Fixed fee agreed in writing first.
Fixed first: one decision-ready view built from existing data, without breaking governance or lineage, with the selection criteria attached to every recommendation.
Fixed first: what each role needed, for doctors, nurses and administrative staff; then more than 300 people trained hands-on, and five AI use cases prioritized.
80%of participating staff demonstrated improved data skills
No. It is how every engagement runs. You see it in what you keep: a scored baseline, a ranked shortlist, a signed Decision Brief, a decision log and a handover your team can run from.
Do we have to go through every stage?
No. The stages are groupings, not a sequence to complete. Most engagements start in one and pull in two more, and security runs through all of them.
Why both stages and phases?
They answer different questions. The stages say what the work is: set direction, build, secure, make it stick. The phases say when it happens inside one engagement. Security and governance are not a phase because they run across all five.
How does the maturity model relate to the stages?
The stages organize the work; the six dimensions measure readiness before and during it. A dimension that scores low points to the one engagement that fixes it, as listed in the scoring sheet above.
How is this different from a maturity survey?
A survey scores your answers. The Assessment Sprint scores the same six dimensions against your systems, with four tests of evidence on each: is it documented, owned, measured and tested.
Do you publish the whole model?
We publish the six AI-readiness dimensions in full. The complete model reaches wider; it is applied in the Assessment Sprint and in our engagements, where it is scored against your systems rather than described in general.
Does it work with frameworks we already use?
Yes. Governance work follows NIST AI RMF and ISO/IEC 42001 where they apply, and security work maps to NIST AI RMF and ISO/IEC 27001. Data IQ™ organizes the work; it does not replace the frameworks your auditors expect.
See Data IQ™ applied to your own systems.
Thirty minutes with the person who would lead the work: where your organization stands on the six dimensions, which stage to start in, and what the first document should decide.
