Strategy

Big Data Fosters Stronger Customer Personalization

Deeper relationships lead to higher sales by helping companies understand customer behavior and the factors that influence purchase decisions.

With the types and volumes of data they collect expanding exponentially, more companies are investing in tools to help them make sense of their data and unlock value hidden within transactional information.

Big data offers promising benefits in a number of areas, but one aspect that’s paying off is automating customer personalization to create deeper relationships that, in turn, lead to higher sales by helping companies understand customer behavior and the factors that influence purchase decisions.

“We process trillions of transactions annually from more than 107 million cards….and $952 billion in billed transactions,” said Swatee Singh, vice president engineering, big data capabilities, at American Express Co., at a meeting of FEI’s Committee on Finance and IT. “But when we take in that data, what can we do with it? That’s where big data comes in.”

Singh said American Express is analyzing its transactional data, in part, to generate promotions and discounts that are personalized for each cardholder based on criteria including his or her purchase history, travel plans, location, purchases by customers with similar interests, and other factors.

“Our transactional data fuels offers that are relevant for our consumers and valuable for our partners,” Singh said. “We can offer personalized benefits through our mobile app, for example, that are personalized based on your location and purchases.”

For instance, Singh said cardholders living in New York who purchase an airline ticket to Phoenix will start receiving, before their trip, personalized emails about restaurants and events at their destination. In Phoenix, they can access offers and information geotagged within a couple of blocks of their location via their mobile devices.

And by examining merchant transaction data, Singh said American Express can discover correlations such as the restaurants or coffee shops cardholders patronize before or after attending a play or going to a movie theatre.

“This personalization provides a benefit to the customer that helps differentiate us from other card issuers,” Singh said. “That creates value that leads to more use of our cards. It also provides value to the merchants because we help connect them with [customers] instead of making mass offers. Most consumers aren’t interested in receiving more offers.”

For merchants, Amex shares aggregated analytical data to help them understand their customers and financial performance in greater detail. For instance, the company provides benchmarking data to help its merchants, most of which are mid-market companies unlikely to invest in analytics on their own, compare themselves to local or regional peers or understand sales segments by customer demographics.

From a technology standpoint, Singh said American Express relies on the open-source data framework Hadoop and other analysis tools hosted on an internal cloud to perform batch analysis of their transactional data on a large scale, while maintaining appropriate data governance and security.

As artificial intelligence and machine learning are developed further over the next few years, Singh said American Express plans to expand its real-time analysis and notification capabilities.

“Security and data management are critical and evolving,” Singh said. “We want business units to be able to leverage this data without having to be developers…and we want big data to help power our business units’ growth initiatives.”