The business glossary defines net revenue as gross sales minus discounts, taxes, and returns. The term has an owner, an approval date, and a steward in finance. The Power BI measure on the CFO's dashboard subtracts discounts and taxes, and returns never come out.

The gap surfaced in a quarterly review, when the dashboard and the general ledger disagreed by the value of a month's returns. The glossary had been right the whole time, and nothing in the estate had read it.

This brief walks one governed definition from the glossary to the measure that computes it, the way MetaKarta handles each step, and ends with what an auditor sees six months later.

Start by finding where the definition is missing

The first question is how many fields in that dashboard have a governed definition at all. In most programs, answering means someone opening the report and checking fields against the glossary by hand.

In MetaKarta, the Power BI report is already harvested into the catalog. Open its object page, choose More Actions and Open Business Glossary, and the report's fields come back with the glossary term that defines each one, linked for review. Fields without a term show up as gaps.

That list is coverage computed from the report itself. The denominator is every field the CFO sees.

Bind the term to the fields that implement it

A glossary term becomes checkable once it's bound to physical fields. Term classification and semantic mapping link the net revenue term to the warehouse columns, the view, and the Power BI measure that compute it.

Semantic lineage then runs both ways. Definition lookup traces from the measure back to the term it claims to implement, and semantic usage traces from the term out to every report and model that depends on it. Turn on Show Term Definitions in the lineage diagram and the glossary definition renders beside the data flow.

The trace also shows why the numbers differ. Follow the measure back through the BI model and the SQL, and the expression at each step is on screen, including the one where returns were never subtracted.

The next change gets caught earlier

Binding also changes what happens the next time someone touches the estate. A data engineer who proposes renaming the returns column in the warehouse runs impact analysis before the change ships, and the trace now reaches the governed term as well as the measures and reports downstream.

A steward who watches the bound warehouse table gets a notification when it changes, so the steward for net revenue hears about the rename from the platform. The divergence that took a quarterly review to surface the first time shows up as a warning before deployment the second time.

Run the correction through review

The fix is a definition change, so it gets treated like one. With workflow enabled on the glossary, an editor creates a draft of the corrected term and starts the review.

A reviewer from finance recommends approval or requests a change, with a comment logged on the transition. An approver approves, and a publisher moves it into the published glossary. Review Changes shows a table of differences between the published definition and the one under review, so the approver signs off on the actual change.

Most users only ever see the published version, which can't be edited directly. Stewards spend their time on the review itself, with the workflow carrying the routing that used to live in their inboxes.

Compile the approved definition into the dashboard

An approved glossary entry is still a description until the BI tool computes it. Ontology Modeling in Semantic Hub binds the glossary term to the governed metric in a semantic model, so the business vocabulary and the calculation stay attached.

Semantic Hub then compiles that metric into the native Power BI semantic model and into a Snowflake semantic view, with each target running its own artifact. The CFO's dashboard computes net revenue the way the glossary defines it, because both read from the same approved definition.

Documentation describes what should be true. Compilation makes it true.

Six months later, the auditor asks

Internal audit samples the revenue figure from the quarter before the correction and asks what net revenue meant at quarter close.

The History tab on the term records every change with its author and timestamp, and the workflow comments show who reviewed and approved the correction. Compare Metadata between the two configuration versions shows the old and new definitions side by side, with the lineage that existed at each point.

The audit answer is a query.

Where the governed definition pays off

The net revenue gap was a small instance of a common condition: a correct glossary definition with no connection to the measure that computes the number. Closing it took four steps on one record: find the unbound fields, bind the term, review the change, and compile the approved definition into the BI tool.

Policy applied where the data is used, versioned where it changed.

MetaKarta does that on one shared metadata repository across Data Governance, Data Catalog, Data Lineage, and Semantic Hub. Get in touch to learn more about how MetaKarta carries a glossary definition into the Power BI measure that computes it.