Capitalizing on Big Data Yields Customer Value

Handled strategically, the mountains of customer information that banks collect today can become a valuable new asset. Indeed, blending structured with unstructured data - such as email, website traffic and social media data - is key to turning data into competitive advantage.

The effective use of relevant data, with the appropriate data architecture and analytical tools, can help banks leverage a deeper view of customers. For example, banks will be able to cross-sell and up-sell by proactively contacting customers based on behavioral triggers and key life stages.

Big data can also open doors to new customers—perhaps enabling personalized pricing based on what’s gleaned about their recent circumstances. And it will help improve customer retention, making it easier to gauge whether a customer is likely to take her business elsewhere.

Challenges ranked high by bank executives include responding to demands for more personalized banking services and addressing customer pricing sensitivities, according to research by Accenture Distribution and Marketing Services.

Leading banks are seeking to enrich their customer profiles by actively collecting internal and external data in real time from multiple sources. The customer-centric business model – which banks are still striving to adopt - will propel the demand for large-volume data collection and processing, similar to what many retailers now consider necessary for developing customer insights. Banks will need to process multiple forms and ever-larger volumes of data to arrive at the detailed, real-time insights that will drive customer value and therefore, business value. At the same time, banks must be mindful of new regulations limiting the use of customer information for marketing purposes.

Bank IT leaders need to recognize, however, that the trend toward multiple data forms and large data amounts is not just about volume. Increasingly, IT leaders’ effectiveness will be measured against their ability to bridge the gap between the structured and unstructured worlds.

We’re not talking about “rip and replace” but rather, taking a balanced approach to technology investments. Thus, banks may continue using standard data management, aggregation and reporting and other traditional technologies for financial and regulatory reporting and other functions not requiring big data. On the other hand, investments in emerging technologies such as Hadoop are needed for applications relying on data discovery approaches, such as credit card fraud.

The key to unlocking the value from big data is to manage it differently than it has been in the past. Big data enables advantages over the conventional approach to assessing and analyzing data.

For example, banks traditionally centralize their data into a single place, structure it, and then analyze it – a lengthy, expensive process with long lags between insights and actions. Consider a card company wanting to integrate a new data feed regarding customer fraud behavior into an existing data base of customer payments behavior. Currently, the company would first have to redesign its data model, integrate the new fraud data, re-aggregate it and finally, do the analysis.

In a big data environment, however, the company wouldn’t be delayed by having to restructure its entire data base. The fraud data could be immediately ingested and analyzed, thereby saving months of down time and quickly enabling new insights. The emerging technologies in a big data environment enable data scientists to conduct data discovery by “bringing the analytics to the data” as opposed to the other way around.

A big data strategy, of course, won’t solve all challenges. The strategy is most appropriate in situations where there is some combination of high data volumes, unstructured data and algorithm complexity combined with the need for quick decision making.

One large North American bank recently transformed itself into a big data analytical player in an effort to obtain a more comprehensive view of its customers and grab a greater share of customers’ wallets while minimizing risk. To that end, it incorporated consumer credit, home ownership, demographics and psychographic data to drive higher mortgage and home equity loan volumes, within the limits of regulations pertaining to customer targeting.

Another bank, based in Europe, combined its customers’ social media profiles, including Twitter accounts, with card spending patterns to sharpen its targeting of customers for cross-selling products and services.

The big data approach is particularly effective when information is needed fast, for example, to capture selling opportunities while a customer surfs a bank’s web site. Offering the customer a specific credit card requires a big data solution to assess his credit risk in near real-time if he is going to make a decision while online. Another application is in the risk area, such as identifying whether that card applicant is a fraud risk. Banks can also apply a big data approach to real-time pricing of a mortgage online that is geared to a particular customer’s profile.

The data discovery process should involve a collaboration of the business and technology teams, as well as the data scientists. To manage costs and hasten data discovery, banks can harness open-source technologies used to process large data sets, such as Hadoop, MapReduce and NoSQL.

As banks transform themselves by becoming more customer-centric, they need to leverage big data technologies to innovatively serve customers and drive business value in the process. Without question, those who are among the first to figure this out will enjoy a competitive edge.

Brian McCarthy is a managing director at Accenture’s North America Financial Services Analytics practice; Wayne Busch is a managing director at Accenture's North America Banking practice.