Advanced Analytics in the New Economy

Customer Experience
Customer Experience

Economic pressures are forcing conversations about the importance of predictive analytics to a place of prominence in organizations. Companies now not only strive to analyze data from traditional channels, but also from their websites, social media, gaming sites, and mobile. Although the rise of interactive media presents its share of challenges for the practitioners of advanced analytics, it also creates opportunities.

Jeff Nicholson, vice president of product marketing at Portrait Software, cites the failing economy as the reason marketing organizations are under pressure to ensure that communications to customers are highly relevant. But moving to a tailored-dialogue communications strategy requires an analytical approach to do it correctly. "The marketing organization is under a magnifying glass and budgets are under pressure," Nicholson says, adding that this is a wake-up call. Marketers are working harder to make sure that when they send messages, those messages are relevant.

While creating relevant messaging is critical, it can be an arduous task. Continuously emerging customer channels mean that the amount of customer data is exploding and producing more data silos than ever before. "The challenge is, you don't know where to start," says Daniel Ziv, vice president of customer interaction analytics at Verint."There's too much information and too many needles in too many different haystacks." But the organizations that are succeeding are taking an intelligent approach to multichannel marketing, gaining a holistic view of the customer, and predicting customer behavior with a higher degree of accuracy.

However, according to Jay Henderson, director of product marketing at Unica, understanding customers in social media and mobile is still in early adoption. Even understanding how customers behave on websites is only now teetering on the edge of being mainstream. "Identifying [customers] across all the touchpoints is a big challenge for marketers," he says, adding that as marketers link customers' Twitter account to their email and Facebook, they will be able to develop a cross-channel view.

Henderson calls this cross-channel user identification-linking a customer's name and address with all the other ways he may identify himself, e.g., connecting his name and physical address to email, and to social media accounts, etc. He cites a variety of techniques to do this, including the simplest one of asking customers what their Twitter handle is and to send the handle in an email. "Connecting the dots across all these ways is a fairly complex task," he says.

A related challenge is analyzing the different types of data from social media. Colin Shearer, senior vice president of strategic analytics for SPSS, an IBM company, says that he sees a trend toward marketers not only analyzing information about customers like past purchases and demographics, but also analyzing their attitudes, behaviors, preferences, and desires. The challenge is analyzing the unstructured data to find what customers are saying about products and understanding the context of what people are saying, who influences others, and what types of postings evoke a response. "Once you've done that, you can be looking at which things you can do to market to them in the future," he says.

The proliferation of data in the future
Unica's Henderson says the data explosion hampers marketers from obtaining a true holistic view of customers. Without that view, predictive analytics is difficult to achieve. "This proliferation of channels will continue moving forward," he says. "Marketers will start to be much more disciplined in how they test these new channels, experiment with them, and nurture them."

And Dave Menninger, vice president of product management and marketing at Vertica, predicts that as data and channels become more complex, organizations will see a need to start hiring statisticians to conduct the modeling. "There is more knowledge that is required today," Menninger says. "You just can't automatically do predictive analytics. It's about the quality of your model. If you don't understand the techniques you're applying, you're not going to get good results."


MNYL Analyzes Customer Channels and Increases Sales
Max New York Life Insurance Company Ltd. aims to be the most admired life insurance company in India. MNYL is a joint venture between Max India Limited, a multi-business corporation in India, and New York Life International, the international arm of New York Life. In line with its vision, MNYL set a goal to grow profitably by retaining and cross-selling to high-value customer segments. The company also wanted to supports its 72,493 agent advisors in its 715 offices with quality data.

As a result the company zeroed in on four critical customer value management programs: share of wallet, persistency, business intelligence, and data quality. "The challenge was to come up with a simple process to align different functions and deal with issues related to converting data to business insights," says Nagaiyan Karthikeyan, head of business intelligence and analytics.

In the past MNYL had conducted broad cross-sell campaign blasts, which accounted for 7 percent of total revenue. Using SAS analytics, MNYL targets the top three customer segments across direct mail, the Web, email, SMS, RSS, and MNYL customer portals. The cross-sell offers are tailored to the different contact scenarios based on customers' value, their propensity to buy, their propensity to pay, and their propensity to lapse.

These efforts have helped MNYL increase the number of new sales from existing customers from 7 percent to more than 20 percent, and improved the premium revenue by nearly 40 percent with shorter sales cycles. With a favorable economy, the company believes it can generate 25 to 30 percent of all new sales from existing customers. "The plan is to now extend our learning to acquiring new prospects that fall within our desirable customer segments, and who yield the highest profitability in the short and long run," Karthikeyan says.

MNYL also analyzes social media for complaints and monitors conversations about the brand. Karthikeyan says that MNYL used the customer interactions in social media to create a new segment called actively involved customers. "These customers are found to have exhibited higher repurchase behavior once their concerns or needs have been addressed successfully," he says.

MNYL intends to increase its cross-sell revenue to 40 percent of overall new business and increase its retention rate in key segments to more than 90 percent, Karthikeyan says. "Since life insurance is a referral business," he says, "we have devised a campaign process of extracting high-value leads from referring segments to further bolster our acquisition of health and wealth."