Organizations today are focused on achieving growth by improving their customer acquisition, cross-sell, and retention activities, as well as by providing better service to increase customer satisfaction. Many of these organizations have made significant investments in customer relationship management (CRM) systems, rebuilding their call centers, branch systems, and Web sites over the past few years.

However, the effect of CRM on these initiatives has been limited. Why is that the case? Because although these CRM systems enable organizations to lower their interaction costs, they don’t indicate what and when to communicate in order to, for example, improve cross-sell rates or increase customer retention.

Predictive analytics enables organizations to gain this necessary insight by predicting individual customer needs and preferences, and providing timely recommendations that can be used to drive more effective marketing campaigns. As a result, it allows organizations to quickly generate increased revenue within their existing CRM environments.

Marcel Holsheimer, Vice President, Marketing, at SPSS, shared with Predictive Analytics Insight some of his views on why predictive analytics is succeeding where CRM has failed, as well his vision on how predictive analytics can take an enterprise to the next level.

PAI: How is predictive analytics helping to generate value in existing CRM systems?
MH: Predictive analytics allows an organization to accomplish three critical things. First, it enables them to understand their customer needs and preferences by building a single view of the customer. Predictive analytics combines customer information from different data sources, but also allows organizations to gather additional information on customers by proactively surveying them to gain new insight. For example, when a customer applies online for a loan, if the organization were to have a better sense of this customer’s objectives (i.e., what is the loan for?) it might be able to tailor its offerings and potentially even cross-sell another product, such as an insurance policy.

Predictive analytics also enables organizations to accurately anticipate, or “predict,” an individual customer's future behavior, product needs, attrition risk, and fraud risk, as well as his preferences regarding the way in which he is approached in term of the tone of voice used, the timing, and through which channel.

And finally, predictive analytics enables organizations to act based on this information by providing an instantaneous recommendation, either directly to the customer (e.g., through the Web or an ATM) or to customer-facing employees such as call center agents. For every customer contact, regardless of the channel, predictive analytics defines the best possible action to take. This could be, for example, making a cross-sell offer for a specific product or a discount offer to a customer whose profile indicates an increased risk of churn.

Typical results for the application of predictive analytics to customer interactions include a 100% increase in direct mail response rates and up to a 50% conversion rate on offers made through call centers.

PAI: Can predictive analytics be applied across multiple channels?
MH: Yes. Predictive analytics is generally applied across multiple channels. Most organizations will start with a single channel—such as the Internet—but gradually extend it into other channels. Some organizations already apply predictive analytics across all channels, including ATMs, call centers, branch offices, the Web, and direct mail. By making offers through various channels consistent, each message reinforces the others, further increasing revenues.

PAI: Does predictive analytics have applications beyond the sales and marketing departments?
MH: Yes. In fact, this is the way in which the most advanced organizations are beginning to apply predictive analytics. They understand that creating customer value is not solely dependent on marketing and sales activity, but is also influenced by any interaction the customer has with the company, such as service interactions. These organizations have begun to apply predictive analytics to these business processes.

Organizations involved in activities such as credit scoring or loan application approval gain significant benefits from the application of predictive analytics. For example, the marketing department of a large U.S. bank uses risk-related predictions to exclude high risk individuals from its loan offer campaign, as they are unlikely to qualify for the loan being offered. And the same is true in reverse; the loan department uses a customer segmentation from the marketing department to assess customer value potential, and then applies this insight to determine whether to approve a loan application.

PAI: What should organizations seeking to successfully implement predictive analytics keep in mind?
MH: Predictive analytic provides an advanced and effective strategy for companies aiming to optimize their customer interactions across a number of different channels. It is, however, essential to always introduce this approach in phases. In this way, the results in each phase can be properly measured and the available customer data can be fully leveraged.