Financial products have grown beyond simple savings and deposit offerings to sophisticated financial instruments, stock trading, investment funds and insurance.
At the same time technological advances have allowed automation of banking processes in the back and front office, as well as self-service customer access through ATMs and Internet banking.
In spite of these developments, banking has become commoditized during this period of change. Many banks offer the same access and delivery channels, products and services to the market. In this process, customer relationships have lost the “personal touch” despite the breadth of “touch points” available to customers from the traditional branch to “mobile” phone banking.
How to differentiate? How to be a “one to one” personal bank while also being a global player? Why should a customer choose one bank over another? How can individual relationships be maintained, satisfied, and grown profitably over time?
Can today's predictive analytics unlock the power of Customer Relationship Management Solutions to provide true “one to one” personal banking on a global scale?
Yes, predictive analytics is the key for banks to unlock valuable insight from their oceans of data about their customers and make better decisions for themselves and their customers. It also enables to leverage and increase ROI from substantial investments already made in operational, data warehouse and business intelligence systems.
Operational systems collect all information about customer transactions. CRM systems collect information about complaints, service requests, inquiries, customer profile, customer surveys and sales related customer interactions. Data warehouses rationalize data from all these systems and Business Intelligence solutions provide enterprise reporting and dashboards for executive management. These systems provide the data used by predictive analytics to provide insight and specific recommendations to truly implement a “one-to-one” approach to customer relationships.
In the last decade, the most advanced banks have used traditional predictive analytics to improve targeting for a few of their largest marketing campaigns. Though traditional predictive analytics require too much time and expertise to be used on all marketing campaigns, it has been used increasingly and stories of success proliferate.
Early predictive CRM success stories involve large statistical staffs and expensive specialty warehouses to develop and maintain predictive models.
• Capital One conducts more than 30,000 experiments a year, with different interest rates, incentives, direct mail packaging, and other variables. Its goal is to maximize the likelihood that potential customers will sign up for credit cards and that they will be able to pay back Capital One.
• At Royal Bank of Canada, increasing customer value through predictive offer management is combined with pricing strategies based on predictions of customer risk and value over time.
• National Australia Bank is well known to use predictive modeling in marketing campaigns and has won multiple awards from the industry, independent analysts and marketing associations for their leading work in and application of CRM analytics.
• Wells-Fargo uses predictive analytics to successfully focus on the breadth of customer relationships, measured in part by the number of Wells-Fargo accounts owned by each customer
• Banks use predictive analytics in real-time for credit risk solutions for credit cards. At the time of charge, transactions are “scored” for credit risk and are either approved or declined, thereby substantially reducing credit card fraud.
It has become clear that predictive analytics brings in competitive advantage in banking and analytics can be the core technology helping banks move from product centric to customer centric operations.
With all these success stories, why haven't predictive analytics been applied more broadly to personalize banking operations? Partly, data was not available or collected for analysis. Also, traditional predictive analytic technology is too slow and labor-intensive to be useful for most business problems. Finally, executives have not fully understood and, therefore, not championed CRM business strategy.
For a CRM strategy to succeed it must involve cultural and business changes and it needs to be a business strategy and not a technology solution. Treacy and Wiersema, in their widely accepted Discipline of Market Leaders, point out that leading businesses have one of three primary competitive advantages that are identified and leveraged.
• Operational excellence
• Product leadership
• Customer intimacy
I suggest that 1 and 2 are required disciplines and some differentiation can be achieved, but the concentration on customer centricity and intimacy will define the bank and achieve loyalty to boost growth and increase shareholder value as a high performance organization.
Customer intimacy is a CRM strategy that requires executive commitment and operational focus. Products must be developed for each market segment and offered to individuals based on their personal profile and history. Performance must be measured in terms of customer satisfaction, acquisition, retention and profitability. Employees should know what to offer to satisfy their customer's needs and help them plan for their financial future, creating a lasting and trusting relationship. Today's new class of predictive analytics can make all of this possible, unlocking the power to provide true “one to one” personal banking on a global scale.
Why is this possible now?
Consulting companies and software vendors are automating specific predictive solutions and providing broad statistical toolsets to empower experts to build sophisticated predictive solutions. KXEN Inc. (headquartered in San Francisco, California USA with research and development based in Paris, France) has developed a new approach to predictive automation. Based upon recent advances in Structured Risk Minimization Mathematical theory (see sidebar), KXEN's Analytic Framework automates many of the most difficult processes in predictive analytics, reducing the risk of error and omission, and dramatically scaling up the number of models that may be built and used. Modeling is very fast, making it possible to model business questions previously not practical, in a timely and cost-effective way.
The components and models can easily be embedded in current business and software processes for bank operations, risk management and marketing. Predictive analytics help experts to be at least ten times more productive, solving more business problems and building “predictive modeling factories”, while business managers can use predictive applications to gain insight, predict results, monitor progress and make better decisions in everyday operations.
This new breed of automated and advanced predictive and descriptive analytical tools are suited for today's fast paced business and can be embedded in enterprise applications. Customer intelligence can now be delivered to the right customer at the time he is actually interacting with the bank, through the delivery channel of choice with consistent advice based on predictions derived from enterprise data.
Banks are embracing these new advanced predictive and descriptive analytics to enhance the ROI on existing in-vestments in Core Banking Systems, Marketing, Financial Systems, Anti Money Laundering Systems, Anti Fraud Systems, to assist in Basel II and Sarbanes Oxley Compliance, Audit Systems, Business Intelligence Systems, Analytical Data Warehousing Systems and particularly Customer Relationship Management Systems.
Software vendors are beginning to use, embed or integrate these robust predictive components in specific banking applications.
Predictive Banking
Today, banks can use predictive technology for every business decision supported by data. Analytic framework enables banks to uncover opportunities, explore trends and make predictions in minutes or hours instead of days or weeks. By adding such a framework to their existing applications, banks can dramatically scale up their predictive analytics to automate enterprise-wide decisions. Major banks are already scaling up predictive analytics for CRM and credit risk analysis.
Banks are also expanding their use of predictive analytics into pricing, human resources, asset management, cash flow predictions, investments, anti-money-laundering, identity theft and other business areas supported by data. Insight from predictive modeling can help executives identify winning strategies, core metrics, and key performance indicators, as well as be alerted to anomalies and competitive and consumer trends.
From marketing optimization to executive dashboards, predictive analytics can unlock the power of information within corporate data for better execution of CRM strategy and beyond!.
Statistical Learning Theory
Vladimir Vapnik is the father of Statistical Learning Theory, which is the foundation of KXEN components. With the concept of the VC (Vapnik- Chervonenkis) dimension, Vapnik introduced a new paradigm for data modeling. Instead of having to make assumptions on the underlying data distribution, this theory leaves data as it is, and allows modeling without having to limit the number of attributes. The resulting algorithms produce robust and high quality models in a fraction of the time it takes with traditional tools. Vapnik showed that many modeling techniques, such as controlled (e.g. weight decay) neural networks, are part of this theory. In a sense, he is the first to have proven in a mathematical way “why” neural networks provide high quality predictions when properly tuned.
Author:
Milovan Puz
Managing Director, Australia / New Zealand
KXEN Inc.
Submitted By:
Ashu Bhardwaj
12129
Finance Stream
No comments:
Post a Comment