Eight questions on whether your data estate can support serious commitments: ownership, quality, access, integration, reporting trust, governance, skills and whether data actually reaches decisions.
Analytics programmes, AI initiatives and system migrations all stand on the same floor: the data estate. This assessment scores whether that floor bears weight, before you find out by loading it. Eight questions, fixed scoring, in-browser.
Last reviewed 3 July 2026 · Free and ungated
The initiative your data estate will carry is easier to challenge before it is funded than after it is live. Selected senior operators from the Global Board provide that challenge in a confidential report.
Pressure-test this decisionAnswer for the data domains your next major initiative actually depends on, not for the estate in general.
The estate cannot yet bear the initiatives being planned on top of it. Anything ambitious (AI, advanced analytics, a migration) will spend most of its budget rediscovering these gaps under a deadline, which is the most expensive way to find them. Unlike a culture problem, this one is concrete and fixable.
Foundations are being laid but unevenly: typically decent pipelines and tooling with soft ownership and governance that meets without deciding. Initiatives succeed here when a strong team drags them through, which reads as capability but is actually dependence on heroics.
The estate is genuinely serviceable: owned domains, measured quality, numbers leaders mostly trust. The residual risks are concentration (skills and knowledge in a few heads) and the gap between operational data discipline and strategic decision usage, which is usually the last thing to mature.
These answers describe a data estate run as infrastructure: owned, measured, governed and actually used. The remaining exposure is subtle: trust. When numbers are believed by default, an error propagates further before anyone questions it, and models built on the estate inherit its blind spots invisibly.
Data programmes attract frameworks the way transformations attract consultancies, and both can run for years without anyone asking the operative question: will this estate carry the specific weight we intend to put on it next? The eight questions here are scoped to that question, which is why the assessment asks you to answer for the domains your next initiative touches, not the estate at large.
Every question has three answers worth 0, 5 or 10 points; the total is expressed out of 100 and lands in one of four bands. Repeatable, in your browser, no interpretation layer. Re-run it when remediation claims to have finished, and let the movement arbitrate.
In practice, data ownership and reporting trust are the leading indicators. Where domains have real owners, quality and governance tend to follow within a few cycles; where leaders bring rival versions of the same number to meetings, every downstream ambition gets discounted. If you only have time to fix two answers, fix those.
Yes, and answer for the exact data the use case consumes. AI initiatives inherit every weakness in their training and input data, and the data-quality and ownership questions here are where most of them are actually decided.
Most land in the developing band, and that is workable: readiness only needs to match ambition. A 45 with a modest analytics roadmap is fine; a 45 underneath an enterprise AI programme is a funded incident.
Use the operational test in the options: where do quality problems surface? If the answer is "in front of an executive, via a wrong number", score low regardless of what the quality tooling claims. The point of discovery is the measure.