Under Credito Valtellinese’s operating structure, innovation projects involving IT and business process optimization are handled by its internal service unit, CSS. To build the analytics foundation for a customer-centric business model, CSS engaged IBM Business Partner SDG Consulting Italia S.p.a. and IBM Interactive Experience and Marketing, a division of IBM® Global Business Services® – Business Consulting Services, to implement a comprehensive customer analysis and multichannel campaign management solution. The analytics support component uses cognitive technology to identify, capture and index unstructured, free-text data stored within the company’s mainframe-based transaction record systems, including information on financial or banking products owned by the customer—such as credit cards, loans, mortgages and insurance—that were not issued by the bank, along with online shopping habits and personal financial arrangements. Using natural language processing (NLP), the solution filters this unstructured text data and creates a detailed profile for each customer, storing it in a repository designed with a customer-centric data structure. The multichannel campaign management component of the solution uses these customer profiles and insights in two key ways. They support market segmentation efforts by analyzing customer transaction patterns to design and test more representative segmentation models, and they target multichannel outbound marketing campaigns. The net functional gain for the marketing department employees is the ability to run sophisticated cognitive analytics on their own—without IT assistance—against a reservoir of valuable unstructured data that had been effectively off limits. Further, this means marketers have essentially no limit to their ability to define the granularity of their target segments.
As conceived, the project created a foundation for the bank to deliver more targeted and effective outbound marketing campaigns to its existing customer base. Since deploying the solution, CSS helped the bank increase the conversion rates for its outbound marketing campaigns by an average of 10 percent, representing an estimated 2 percent of net-new revenue growth. The solution also increased the agility and efficiency of the marketing department by making it more self-sufficient in performing analytics against customer data.
The Credito Valtellinese solution is game-changing because it uses NLP technology to identify valuable pieces of information within large volumes of unstructured customer data that were impossible to isolate and extract before. This enabled the bank to vastly increase the precision of its market segmentation and targeting efforts, with commensurate improvements in conversion rates.
Prior to the project, the marketing department was fully dependent on the bank’s internal IT organization to run analytics against its customer accounts databases. In addition to being a slow process, the scattered, product-centric orientation of the data made it next to impossible to extract deep insights that could be put to work in marketing campaigns. The new cognitive solution essentially breaks the barrier to identifying insights by using NLP to identify and index the data in a way that supports customer-centric marketing.
The solution employs both structured and unstructured data from internal sources within the bank, principally the company’s core banking systems and customer account activity databases.