Business problem

Finance, sales and the board each reported different revenue numbers, so monthly reviews were spent reconciling spreadsheets instead of making decisions.

Dataset

Billing, CRM opportunity and budget data for 13 rolling months across 6 business units, refreshed nightly from Azure SQL.

Methodology

• Agreed a single certified metric layer with finance.
• Built a star schema with incremental refresh over fact tables.
• Implemented row-level security by business unit.
• Added field parameters for flexible executive drill-down.

Key findings

• Two business units were double-counting renewal revenue.
• Pipeline coverage below 3.0x reliably preceded a missed quarter.
• Margin erosion concentrated in one discounted product line.

Recommendations

• Adopt the certified model as the sole source for board reporting.
• Trigger pipeline alerts when coverage drops under 3.0x.
• Review discount policy on the underperforming product line.