What Navy Federal found in its synthetic data pilot

Navy Federal Credit Union recently completed a pilot program testing the use of synthetic data—artificially generated information that mimics real consumer behavior. For retailers and lenders, this technology promises a faster way to test new financing products or app interfaces without risking sensitive customer data or waiting months for human surveys. The pilot showed that AI-generated data is highly effective for testing technical workflows and predicting basic transactional behaviors. It allows companies to run 'what-if' scenarios on credit applications or payment flows at a fraction of the traditional cost. However, the study also revealed a major limitation: the 'human element.' While the synthetic data accurately predicted how users might navigate a menu, it failed to capture the nuanced emotions and complex motivations that drive actual borrowing decisions. For business owners, this means that while AI can help you refine the mechanics of your financing program, you still need real feedback from your customers to understand why they are choosing—or rejecting—your payment options. Use synthetic data to fix your digital bottlenecks, but don't fire your customer research team just yet. The technology is a tool for efficiency, not a total replacement for understanding the consumer psyche.

Source: Banking Dive

← Return to the library· Submit a correction