Hyundai Card is classifying and structuring customer data that had been listed without meaning, maximizing synergy with artificial intelligence (AI). Because AI requires proper data to deliver results, the company laid the groundwork for this. In the market, there is an assessment that Hyundai Card's foresight, which has driven large-scale investment for more than a decade by emphasizing the importance of data, is now paying off.
According to the financial industry on Friday, Hyundai Card is applying and using various AI models, including "Mother AI" and "Allocation AI," to its in-house developed AI platform "UNIVERSE."
UNIVERSE is a platform that analyzes customer data with AI and can be used for various tasks such as marketing and risk management. Mother AI combines diverse consumption data to identify customer groups highly likely to respond to a particular marketing campaign. For example, in the case of online food mall marketing, an employee would combine relatively familiar conditions such as "customers with children," "customers with extensive online payment experience," and "customers who use marts on weekdays." Mother AI, on the other hand, combines even data that appears to have low relevance, such as auto finance usage characteristics and travel spending patterns. This was found to produce a response rate up to six times higher than conventional methods.
Allocation AI calculates how to distribute limited marketing resources such as coupons and points for the greatest effect. Assuming a situation where only one coupon among gas, beauty, and coffee coupons must be provided to one million customers, Mother AI first identifies the customer groups highly likely to respond to each coupon, and Allocation AI optimizes the distribution quantity. Through this, the marketing design period, which had previously taken as long as three days, was shortened to 20 to 40 minutes.
Behind these achievements is Hyundai Card's "data structuring" strategy, which it has pursued for more than a decade. After declaring "Digital Hyundai Card" in 2015, Hyundai Card invested a total of more than 1 trillion won in the AI and data science fields. Of this, 500 billion won was put into data structuring. The company devoted as much effort to building a data system that AI can read and judge as to developing AI models.
The result is Hyundai Card's "PDI tag" system. It converts vast amounts of data, such as card payment records and location and time, into an intuitive form that shows customers' tendencies and behavioral characteristics. While the "P (Primitive)" tag is basic data derived from actual behavior, such as shopping areas and payment time zones, the "D (Derived)" tag is an index that estimates customers' tendencies or future behavior possibilities based on this. The "I (Inferred)" tag is used to find target customer groups by combining several D tags to fit a specific marketing purpose.
The full investment in data structuring reflected the strong will of Hyundai Card Vice Chairman Chung Tae-young. Chung believed that the conventional market segmentation method of dividing customers by gender, age, and occupation makes it difficult to keep up with increasingly diverse tastes and lifestyles. His judgment was that "hyper-personalization," which precisely interprets the payment data held by a card company to provide services tailored to each individual customer, will become a core competitiveness of financial firms going forward. Global software company Palantir alone is securing competitiveness in corporate and defense projects by leveraging "ontology," which structures data into a form that AI can understand.
This strategy by Hyundai Card led to growth in business performance. Hyundai Card's membership grew from 11.04 million in 2022 to 12.67 million last year. Over the same period, its personal credit sales market share expanded from 16% to 17.5%. "The fact that we solidly established a data science-based business strategy served as a catalyst that sharply grew our market share and membership," a Hyundai Card official said.
