The European bank sought a text analytics solution for their event marketing campaigns that would enable them to extend product offers in response to a customer’s real-life events. Ataccama delivered a solution that brought a tenfold increase in campaign success.
While event-driven marketing in connection with structured data analysis (i.e. data organized into spreadsheets or bank forms) has been popular for some time, recent advances in big data capabilities have made it possible to pursue text analytics even with unstructured data. This includes customer feedback forms, social media posts, client emails, and more.
For the bank, Ataccama was able to use text analytics and Ataccama Big Data Engine (BDE) on unstructured data to uncover new information about bank clients and successfully offer these clients products tailored to recent life events.
Ataccama Big Data Engine (BDE) is a powerful technology built for high performance, scalability, and the rapid processing of huge volumes of data. BDE covers the entire data integration and management process, including data extraction, import to Hadoop, preparation, cleansing, and general processing within Hadoop.
Working with three types of data (bank clerk notes, IB transaction descriptions, and card transaction identifiers), we identified two specific client groups as ideal recipients for a new product offer from the bank. The first group consisted of parents struggling with financial solvency at the start of the school year, and the second group was a pool of people planning to go on vacation in the near future.
Previously, the bank had no way to determine if a client had a child, unless they already owned a child-related bank product. With text analytics, we were able to use keyword analysis to identify which clients had children, how many they had, and which ages their children were. The bank then offered these parents a short term loan that coincided with the start of the school year.
Using text analytics to parse references to hotel reservations, plane tickets, travel insurance, and related topics, another group of customers was identified as planning an upcoming vacation. The bank was able to extend these clients an offer for a new credit card for any unplanned expenses that might arise while traveling.
We reached a total conversion of 6–7%, or a 2x increase over the traditional combination of direct mail and call-center calls. The estimated selling potential concerning half of this group (obtained using the traditional model) was about five for every thousand people. The bank was thus able to increase the success rate of its sales more than tenfold.
When it comes to event-driven marketing, timing is a crucial factor in a successful campaign. The process of transforming unstructured data was quick, allowing the system to identify thousands of relevant clients on a monthly basis. With the use of text analytics over given intervals, we were able to identify relevant customers in a timely way and substantially increase the success rate of sales.
Our solution can be easily modified and applied across a range of fields, from marketing to risk detection, collections, and more. Get in touch with us to learn more about text analytics.