Western Union is a company for which management of data is central to its operations, enabling its customers to transfer money to and from almost every nation on earth.
Since setting up shop in the US over 160 years ago, the firm has grown its business to serve 70 million customers in 200 countries, enabling 650 million transactions every year. This has contributed to what has been described as "one of the world's largest enterprise data sets", with masses of information generated across its consumer and B2B operations.
Capturing and analysing this information can be a challenge, as Western Union shifts its focus from cash to digital transactions, but the insight gained is key to the company's primary objective - improving customer experience in an increasingly competitive money transfer market. With the introduction of a host of new firms offering remittance services - from start-ups to larger players such as Vodafone, Wal-Mart and possibly even Facebook in future - Western Union face pressure to modernise its business and provide a widening range of online and mobile services.
Leading the company's data strategy is CIO and EVP of global operations David Thompson. Thompson joined Western Union from security firm Symantec in 2012, where he served as global CIO, and has a strong roots in the technology industry having also worked as CIO at Oracle and PeopleSoft.
Along with CTO Sanjay Saraf, recruited in 2013 to drive digitisation and revamp the firm's technology back-end, Thompson has led a programme of change with investment in advanced analytics tools.
"When I started my first question was: how are we using this data, and what business areas are using this to make decisions about pricing, product, service, availability," Thompson said.
"One of the things I found was that the business was not always leveraging that data as effectively as it could have been.
"Big data analytics have given us an excellent tool because we have now have access to data which we couldn't have gotten our arms around before. In this new world of Hadoop technology and cloud-based analytics solutions, you can get answers in a much more economical fashion, and with dramatic speed."
Big data analytics
Thompson and Saraf have overseen investment in a range of analytics and data management tools in the past two years, building on its existing Oracle databases and Business Objects data warehouse. These tools provided summarised reporting and statistics, but did not enable the level of detail required to build useful customer profiles.
A major component of its analytics plan was an implementation of Hadoop with close help from Cloudera, with the company setting up a 32 node cluster.
"With the size and scale of data we have, the best choice was clearly going with a Hadoop type of ecosystem. This allowed us to start leveraging the scale of structured and unstructured data to drive decision making," said Saraf. "We did it at a breakneck speed, and we stood it up in five months."
Sign up for Computerworld eNewsletters.