Understanding Financial Data and the Value of Analysis

The core of data analysis lies in defining key metrics suited to the characteristics of each domain and using them to advance services. While shopping data focuses on payment patterns and advertising data focuses on impressions and clicks, financial data focuses on understanding the changes and scale of transaction history.

Typical app log data records users' content consumption and behavior (clicks, impressions, etc.) and is generated at the moment users use the app. In contrast, financial data differs in that it is continuously generated even when users are not using the app, through inter-account transfers, automatic transfers, cashback, and so on.

Financial data has high complexity for the following reasons.

• Account-centered structure: It is based on various types of accounts—such as deposits/withdrawals, savings/installment savings, and loans—and their status values.
• Diverse transaction types: Transfers, withdrawals, automatic transfers (firm banking, giro, CMS, etc.) vary widely in purpose and method.
• Continuous occurrence: Data is generated through interactions between financial means even without a user's direct action.

Therefore, to maximize the value of financial data, a process is needed to internalize data understanding, share key metrics, and raise company-wide data capabilities through the democratization of data analysis.