Big data gives KB Kookmin Card a $600m revenue lift

KB Kookmin Card enjoyed a W700 billion ($600 million) year-on-year revenue boost in 2016, which is partly attributed to its new digital marketing strategy.

Jaehoon Park, head of big-data strategy at the card company, said the firm’s platform, powered by technology integration vendor Tibco, has given KB Kookmin Card a multi-year leap ahead of its major competitors in Korea.

Intense competition in Korea’s credit-card industry has saddled the industry with a no-growth situation. The market is saturated, with 80% of family expenditures now paid by cards. The average Korean holds multiple cards, and uses them for increasingly small transactions: 42% of card use goes to purchases below $9, Park says.

Yet there are 22 card companies in Korea, and they are in a constant state of trying to innovate or differentiate. Advances tend to be short-lived; within a few years, everyone will have copied the best ideas.

Park’s team was looking for a way to raise the bar high enough that it would take the likes of Hyundai Card and Samsung Card much longer to catch up. It has done so by adopting a marketing strategy based on real-time data, personalization and better analytics.

The traditional marketing strategy was to email or text customers selected from the company’s database about a campaign, but these were generic messages and eventually people came to see it as spam.

KB Kookmin Card worked with California-based Tibco starting in 2013 to overhaul how it engages with customers. It first introduced a marketing platform using real-time marketing tools to identify specific people who might want a particular offer, based on their other buying or online behavior, Park says. Big-data tools scoured customer data to create more personalized offerings. And it deployed visualization technology to get a clearer view on how customers responded.

Then in 2016 came two advances. It scaled out the architecture with new hardware and was able to run 300 campaign scenarios, a tenfold increase over the usual 30-odd campaigns that card companies handle annually. A third project, toward the end of 2016, applied machine learning to automate specific offers, based on customer behavior.

“Now when customers use our cards, we can analyze where and why, in real time,” Park said, speaking at a conference organized by Tibco. “And we can respond in real time.”

By leveraging Tibco’s systems, KB Kookmin Card minimizes marketers’ dependence on its own I.T. team. “I.T. can be an obstacle when marketers want to try something different,” Park said. Now, however, the marketing team can design a campaign autonomously, and try to implement all kinds of ideas. “It maximizes our creativity and minimizes the program coding.”

The instant feedback also has gotten marketers more interested in client data, he added.

The company uses two Tibco products to implement this new marketing strategy. It uses Spotfire, a software for data analytics and visualization, to investigate customers’ historical data in order to come up with campaign ideas. Then to personalize the results, it uses BusinessEvents, a big-data processing application that applies rules to monitor streaming data. Spotfire is then used to monitor the results and incorporate that into the customer data history, so that the process can be updated and repeated.

“This is not something our competitors can imitate quickly or easily,” Park said.

Michael Connell, Tibco’s chief analytics officer in Palo Alto, told DigFin, “This use of Spotfire at the front end to develop rules about customer behavior, and then embed it with actual actions, and then turn around and do it again – this is what many financial service companies want to do.”

Asked by DigFin what’s next, Park says his focus is on personalization.

Connell said omni-channel retail banking is “kind of the holy grail”. Banks’ complexity makes it hard to enable them to sell any kind of product through any kind of channel, at a price that makes sense. So is identifying the most important elements to get a full picture of a customer. Banks are also working on how to embed artificial intelligence so that sales and marketing people can deploy it in real time.

Park seemed ready to keep working toward that goal: “We want to continue innovation; we want more ideas and solutions.”

This article originally appeared on March 29, 2017.