Improve data quality before adding another dashboard

When teams disagree about a number, the problem may not be the visualization. Different definitions, incomplete records, and unclear ownership can produce conflicting reports from the same business activity. A useful dashboard begins with confidence in the meaning and quality of its inputs.

Define the decision the metric supports

Ask who will use the number and what they will do differently when it changes. Write down the inclusion rules, time window, and source of truth. For example, “active customer” can mean a current contract, a recent purchase, or a product login. Those are different concepts and should not share an unexplained label.

Trace a sample back to the source

Choose a few records represented in the report and follow their path through collection, transformation, and aggregation. Look for duplicate identities, missing dates, inconsistent status values, and late updates. Document where an error should be corrected so that the fix improves the source rather than only one exported spreadsheet.

Assign ownership and useful checks

Every critical dataset needs someone accountable for its definition and handling of exceptions. Add checks that reflect the business rules: missing identifiers, impossible date sequences, or unexplained changes in volume. Make the result visible to the team that can act on it, with a practical correction workflow.

A practical next step

Pick one disputed metric and agree on a written definition. Validate a small sample with the business owner, then add a focused quality check before redesigning the dashboard around it.