Why structured risk minimization is the next trend in predictive analytics.
Posted Nov 15, 2004
Today, just about every major company has spent a big part of its budget to implement a CRM system. As customers we have been promised to receive better service, faster. As marketers we are supposed to have acquired intimate knowledge about the behavior, the preferences, the needs, and the wants of our customers. All of us who are touched by marketing, either on the consumer or on the vendor side, have had ample opportunity to experience what these systems can and cannot do for us.
So what has changed? If my bank is willing to spend millions of dollars to better understand my needs, why does my ATM still ask me if I speak Chinese, Portuguese, or Greek? Why doesn't it know that I always withdraw $60 on Thursday evening? Plenty of data is available for vendors to answer these questions, get closer to their customers, and provide meaningful service, but most companies are overwhelmed by that task.
And when companies do use reporting tools, all they produce are reports of activities that have occurred in the past. The critical piece that's missing is the predictive information about future behavior. For example, my credit card company knows that I have had dinner in Frankfurt, London, and Paris. What the report doesn't say is if these were business trips or vacations. The company knows if I will need extra cash in France and how it can help me finance a new timeshare condo in Cannes.
A Russian scientist has conceived a breakthrough mathematical theory, known as statistical learning theory. A concept of this theory, called structured risk minimization (SRM), is the foundation of a new type of predictive analytic software that automatically analyzes large corporate databases. SRM is on its way to revolutionize the use of analytical software, because it uses sophisticated mathematics to drastically simplify customer analysis. A side effect of this new approach is that it creates predictive models using all the available attributes and variables where previously only a handful found their way into the analysis. Most important, it gives a view of the future based on the past, not just twenty-twenty hindsight.
SRM software can be embedded at every touch point of a customer. It provides the marketer with an automatic tool to anticipate customer needs. SRM-based, targeted segmentation allows the user to identify and address very small customer segments with a relevant product offering. So, a call center representative can have meaningful interactions with a customer, providing better quality customer service. eCommerce vendors can anticipate the next click of a prospective customer on the Web in real time.
Companies like AC Nielsen, JP Morgan Chase, Washington Mutual, and Fidelity as well as industry leaders AT&T, MCI, and Orange have discovered that SRM provides them with a distinct competitive advantage.
CRM systems are here to stay. The next big wave is challenging owners of these systems to make use of their data to predict action, rather than just reviewing what has happened. SRM predictive analytics are on the forefront of this wave, and help companies achieve what CRM systems so far have failed to deliver--true insight and a deeper understanding of their customers. Now then, I am off to my ATM around the corner, anxiously hoping to see the results.
About the Author
Joerg Rathenberg is the vice president of marketing and alliances at KXEN (www.kxen.com), a provider of predictive and descriptive analytic software that can easily be embedded into existing business processes and applications, and the first to have implemented SRM. He has focused on the development and commercialization of emerging software technologies worldwide. He can be reached at firstname.lastname@example.org
Strategies to reduce operating costs and protect revenues.
Sponsored By: Marketo and Real Magnet
Sponsored By: Jacada, Avaya, Confirmit, inMoment and BoldChat
Sponsored By: Genesys, Avaya, Verint, and Aspect
Sponsored By: Informatica