Reporting & Dashboard Audit

What's included

Data sources

• Connector health across every source feeding your reports, including quota limits and sampling
• Blended data review, since Looker Studio joins default in ways that quietly drop rows
• Calculated field logic checked against the raw data
• Date range and comparison period handling, including the reports that break at month boundaries
• Refresh cadence, caching behavior, and stale data
• API quota usage and the reports failing silently under load

Metric definitions

• Metric dictionary showing how each number is defined and where definitions disagree across reports
• Reconciliation between dashboard figures and the source platforms
• Attribution model consistency, since one report on data-driven and another on last click will never agree
• Currency, timezone, and fiscal calendar settings
• Filters applied at page, chart, and source level, including the ones a former contractor left behind

Structure and usability

• Report inventory covering what exists, what gets opened, and what can be retired
• Client-facing report review for clarity, load time, and the questions it leaves unanswered
• Template review if you run the same report across many accounts
• Access control and sharing settings
• Naming conventions and version sprawl

Deliverable

• List of numbers currently wrong, with the corrected figures
• Fixes implemented on the reports clients actually see
• Documented metric definitions you can reuse across accounts

Why it matters

Reporting errors are the worst kind because nobody questions them. A blend drops the rows without a match, the conversion number comes out low, and a client makes budget decisions on it for eight months. Every chart renders fine. Nothing looks broken.

Templates make it worse at agency scale. One bad calculated field replicated across forty client reports is forty wrong dashboards, and you find out when a client cross-checks against the ad platform. Better to find it first.