Bank transaction data

Bank transaction data is categorized, de-identified account activity: income, spending, transfers and bills. Lenders use it to understand affordability and cash flow patterns, fintechs to build categorization and personal finance features, investors to track spending across payment methods, and economists to study household finances.

Typical fields

• Transaction date and amount
• Category and sub-category, such as groceries or rent
• Normalized merchant or counterparty
• Transaction type: card, direct debit, transfer or cash withdrawal
• Account type and region, with a pseudonymous account ID
• Aggregated metrics such as median spend per category

Sources and collection

Data comes from banks or open banking aggregators that receive transactions with the account holder's permission for a service, and only where their terms and the law allow secondary use in de-identified form. Providers remove names, account numbers and other identifiers, normalize merchant descriptors and categorize transactions with rules and machine learning. Many datasets are sold only as aggregates.

Delivery and pricing

Delivery is usually aggregated time series by category and region, or pseudonymized transaction-level panels for modelling, via files or cloud shares. Some providers sell categorization as an API instead of data. Price depends on panel size, granularity, history and intended use.

Before you buy

• Ask about the legal basis and permissions under which the data was shared.
• Confirm account numbers and personal identifiers are removed.
• Check categorization accuracy for the categories you care about.
• Ask how the panel is distributed by bank, age and region.
• Check minimum aggregation thresholds.