Data Warehousing for Fintech: Unit Economics, Cohorts, and the Metrics VCs Track

A Series C fintech prepares for its board meeting. The data team produces the investor metrics: CAC, LTV, payback period, NDR, GRR, monthly burn, and unit economics by customer segment. The CFO reviews the deck and finds that the LTV calculation uses a different churn definition than what was reported last quarter, the CAC includes some marketing costs but not others, the NDR calculation doesn't match the revenue recognition in the financial statements, and the unit economics exclude certain cost categories that should be allocated. The board sees different numbers every quarter not because the business is changing but because the definitions keep shifting. Investor trust erodes. The next fundraise gets harder because the data room metrics don't tell a consistent story.

Fintech data warehousing done right locks the metric definitions first and builds the models around them. One LTV calculation with a documented churn definition, discount rate, and gross margin assumption. One CAC that includes all acquisition costs, consistently, every period. One NDR/GRR that reconciles to recognized revenue. Unit economics that allocate costs consistently using documented methodology. All sourced from a dimensional model that reconciles to the ledger for financial metrics and to the product analytics system for usage metrics. With definitions locked, the board sees a consistent story. With definitions shifting, every board meeting starts with explaining why the numbers changed.