BI Governance
Keeping business intelligence trustworthy — through certified definitions and a semantic layer, without preventing analysts from working.
Definition
BI governance is the set of controls that keep reported numbers consistent and trustworthy while allowing exploratory analysis.
The failure it prevents
Two dashboards report different revenue for the same month, both are defensible, and trust in the whole platform collapses. Recovering that trust costs far more than the governance would have.
The cause is almost never technical: it is that "revenue" was defined independently in two places, with different treatment of returns, discounts, cancellations, or the date used.
The mechanisms that work
A certified metric definition with a named owner. One definition of each key metric, stating inclusions, exclusions, the date basis and the timezone. Owned by a person, not a committee.
A semantic layer that implements it. The definition exists as versioned, tested code that every dashboard and query consumes. This is what makes the definition binding rather than advisory — if the only way to get the number is to call the shared definition, divergence becomes impossible rather than discouraged.
Certified versus exploratory tiers, clearly labelled. Analysts must be free to compute things their own way; the failure is an exploratory number reaching a board pack unnoticed. Labelling addresses that without slowing anyone down.
Lineage. When two numbers differ, "where did each come from" must have a mechanical answer in minutes.
Conformed dimensions and declared grain, because fan-out double-counting is the most common technical cause of disagreement and it produces plausible-looking wrong numbers.
What does not work
A review board approving new metrics, which adds friction proportional to volume and produces shadow spreadsheets rather than compliance.
A catalogue documenting both definitions, which has documented the problem.
Restricting analyst access, which relocates the analysis into tools with no governance at all.
The balance
The failure mode at both extremes is real: no governance produces an untrusted swamp; heavy governance produces shadow pipelines built to avoid it. Target the lightest mechanism that makes the right thing easier than the wrong thing.
Interview question
"Two dashboards disagree about revenue. What mechanism would have prevented it, and what would not?"