Eight versions of "active customer". None of them right.
Every team brings its own number to the meeting. Decisions stall while people argue over whose is right. Trust drains faster than your data lake fills.
Eight definitions of "active customer". Three different revenue numbers for the same week. Source-of-truth debates kill 20% of every leadership meeting. Fix it with a Microsoft Purview governance framework. First trusted KPI live in four weeks.
If half of these sound familiar, this is your fight.
No formal assessment needed, just an honest look at the daily friction.
Three reports give three different revenue figures for the same week.
Finance overrides the BI team's numbers in Excel, just to feel safe.
Nobody can answer "who owns this metric?" in under a minute.
Master data is maintained in three systems and one shared spreadsheet.
Source-of-truth debates eat 20% of every leadership meeting.
Auditors flag inconsistent definitions and no documented lineage.
Why it happens
No clear ownership
Definitions live in heads, not in a registry. Stewards exist on paper, not in calendars.
Fragmented systems
CRM, ERP, marketing tools and Excel each have their own version of customer, product, transaction.
No lineage, no audit
Numbers travel through pipelines you can't trace. When something breaks, root cause is a guess.
What it costs you
Slower decisions
Leadership debates definitions instead of strategy. Every meeting reopens the same numbers fight.
Compliance risk
GDPR and the AI Act demand documented lineage. Without it, every audit is a sprint.
Stalled AI
Models trained on bad data ship bad answers. Trust collapses, and AI initiatives quietly die.
The shortest path from problem to results.
Microsoft-first stack. Belgian and Estonian engineering. Senior team kickoff through delivery to support.
Governance framework on Microsoft Purview
Stewards, classifications, retention and lineage: live in your tenant within weeks.
See solutionMaster data strategy
Where master data belongs in your architecture: one canonical record per customer, supplier, product, with a change workflow that holds.
See solutionTrusted semantic model in Power BI
Certified definitions, row-level security, calculation groups: the spine of self-service.
See solutionWhen the eight definitions of "active customer" became one.
One certified semantic model, audit trail end-to-end
A €200M services group with eight definitions of active customer across CRM, ERP and finance. We built a Microsoft Purview registry, a Fabric semantic model and assigned stewards by domain. First trusted KPI live in week four. AI Act risk register live in week ten.
Data you can't trust is fixable. Let's start with one definition.
Free 60-minute trust assessment: pick the metric that hurts most.
Common questions, direct answers.
How do we get to one definition when we have eight today?
Pick one metric that hurts most (usually customer or revenue), assign a domain steward, document the definition in Microsoft Purview, retire the seven duplicates over six weeks. Repeat for the next metric.
Do we need Microsoft Purview specifically?
No, but it integrates natively with Power BI, Fabric and your M365 tenant, which means lower TCO than Collibra or Informatica. We use what your stack already pays for.
Can we do master data management without buying an MDM tool?
For the first 80%, yes, using Dataverse + Power Platform for the steward workflow, plus Purview for classification. Dedicated MDM tools only earn their keep above ~3M master records.
How long until governance shows business value?
First trusted certified KPI in four weeks. AI Act risk register in ten. Full lineage map in twelve. Beyond that it's continuous improvement, not a project that "finishes".
What happens if our team resists governance?
They will. The fix isn't process. It's making the right thing the easy thing. Stewards spend hours per week, not days. Self-service users get Purview-classified data through Power BI without thinking about it.