The Whole Picture

A multi-channel retailer that can’t integrate data across channels is like the proverbial six blind men trying to describe an elephant. While they might be able to describe the tail, leg or trunk, they aren’t able to give a complete and accurate picture of the animal.

The same holds true for multi-channel retailers who mine data from different channels but don’t bring it together to get a broad understanding of their customers. They may develop an accurate picture of a customer’s behavior in one channel, but they won’t have a complete profile. And they will find it hard to wring the maximum value out of that customer.

The value integrated multi-channel data offer in the crafting of marketing strategies can be seen in the success some marketers have had in using that approach. For instance, analysis of consumers’ cross-channel preferences led Blockbuster Inc.’s Blockbuster Online to offer subscribers a coupon for four free offline movie rentals per month, up from two per month.

The result: 40% of the program’s coupons are redeemed every month and 70% of online customers have used coupons at some point, drawing online customers into Blockbuster stores, says Shane Evangelist, senior vice president and general manager. That additional traffic generates sales of movies, games, candy, magazines and other products not sold online.

“At the end of the day, what consumers are doing is renting movies from us, whether they’re getting them from the store or they’re getting them through the mail,” Evangelist says. “We like the combination of having the consumer feel there is more benefit to the service as well as getting more traffic into our stores.”

More engagement, more value

NFLShop.com has had similar success in using multi-channel data to increase the number of customers using more than one NFL channel—and buying more NFL products. The NFL aggregates customer information drawn from multiple sources, including NFL co-branded credit card programs and a strategic partnership with Sports Illustrated magazine in which the NFL swaps names with the publisher.

“We’re trying to build a direct relationship with our fans,” says Brian Fitzgerald, manager of database marketing for NFLShop.com, the National Football League’s e-commerce site. “The more we engage them in multiple channels, the more value we see.”

Indeed, studies show that multi-channel shoppers are more valuable because they are more likely to respond to cross-selling and upselling. And they represent a growing segment of online customers—55% of U.S. online consumers have shopped across channels, an 8% increase from 2004, according to Forrester Research.

In addition, shoppers who crossed from the web to stores spent more than $125 billion in brick-and-mortar stores in 2005—a 23% increase from 2004, Forrester says. Upselling and cross-selling accounted for about $16 billion of that amount.

The challenge

Those numbers may be enticing to retailers, but they are not easy to achieve. “The challenge for a multi-channel retailer is to understand which channel the customer prefers and which channel is most profitable per unique customer,” says Leslie Ament, director of customer intelligence research for the Aberdeen Group, a research and consulting firm.

And that requires getting the data in order—“no small feat,” Ament says. “That could be a 12- to 18-month project for some companies, depending on how many legacy databases need to be integrated and cleansed.”

With mergers and acquisitions, many retail organizations end up with multiple databases, Ament says. “There could be 12 or 20 legacy database systems holding customer information at one company,” she says. “Or large companies could have different brands and each brand might be in a separate business unit and each business unit has separate customer data.”

Retailers need to weed out redundant listings, bad addresses and other poor quality data. If they don’t, subsequent analysis will result in the classic “garbage in-garbage out.” Once data is consolidated, standardized and verified, the retailer can begin capturing crucial information about which channels a customer prefers, another daunting task, Ament says.

Often a customer will access multiple channels before buying an item, making it difficult for the retailer to determine which channel played the key role in the customer’s purchase decision. Ament notes that she often looks at an outfit in a catalog but will go to the store to try it on and then will buy it online because the color she prefers is only available at the web site. “If you’re a marketer, how do you know what influenced my purchasing decision and to what extent I’m going to interact with you as a customer?” she says.

But technology is evolving that can help retailers analyze multi-channel data, including operational business intelligence and predictive analytics, Ament says.

Different strokes

Operational business intelligence involves using transaction information to see what the customer has bought in the past to make assumptions on what he might buy in the future. Predictive analytics is more complex, involving account statistical modeling based on large data pools within an organization, Ament says.

IRCE 2008 Conference Multi-Media CD-ROM

“A clothing retailer might have different statistical models from a consumer electronics retailer,” she says. “But the philosophy and functionality behind predictive analytics is determining a customer’s propensity to buy and through which channel and at which price point.”

Multi-channel retailers also can find out customers’ preferences across channels by simply asking. That’s the approach Blockbuster took in deciding whether to up the number of movies online customers could obtain at stores as part of their subscription, and it’s the approach it takes before making any offers, Evangelist says.

Blockbuster uses focus groups, consumer surveys, and control group testing to determine which offers have the most likelihood of success, Evangelist says. “We basically check a number of propositions with both the existing customers as well as people who are not using our service today,” he says. “It’s penny-wise and pound-foolish to go to market with an idea that you think consumers are going to like prior to actually talking to consumers.”

Blockbuster also uses an internally developed system to monitor in-store and online traffic and consumer demand for different movie titles. Based on regression analysis of that information, Blockbuster will tailor store inventory to the rental patterns of a specific neighborhood or decide which titles to offer online, he says.

NFLShop.com takes a more complex approach to data integration, drawing data from numerous sources, including the NFL web site and the 32 NFL teams’ sites. A data management division of credit card issuer MBNA Corp. hosts and manages the NFL’s database, which holds 18 million unique names acquired through registrations for everything from the shopping site to NFL fan newsletters, Fitzgerald says. MBNA’s data management division operates independently of its card issuing division, which issues the NFL’s Extra Points co-branded credit card.

The NFL then analyzes the data to determine what types of offers to make to customers, Fitzgerald says.

Maximizing the value

“The more sources you’re in, the more avid and more intimate a relationship and the better potential prospect you are if we send you some sort of targeted message,” he says. “That’s maximizing the value of that relationship. The fans get what they want—more contact, more information, more touch points with the league. And we maximize in building out a relationship with our fans.”

Individual teams have access to the database for information related to their own operations. The information can be broken down by such segments as channel, favorite team, and e-mail address and preferences, Fitzgerald says.

If the NFL wants to further segment its customer base, MBNA overlays demographic information purchased from a third-party provider onto NFL-sourced data. For example, the NFL might want demographic data that would give it insight into women with annual incomes of $70,000 who have made a purchase at NFLShop.com in the last two years.

“That helps us understand our fan base in a more comprehensive way than we historically ever had,” he says. “It gives us a better understanding of who our fans are, what they’re interested in and how they want us to market to them and what they want us to market to them.”

Demographic information providers collect data on about 80 variables, including address, gender, income, education levels and geographic locations.

Fitzgerald won’t discuss the costs of running the data integration program, saying the information is proprietary. “There are additional costs involved, but we’re building it out and we’re growing it out in a positive way,” he says.

Six-fold increase

And while he won’t discuss conversion rates, Fitzgerald says the program has shown promise for achieving the NFL’s goal of getting shoppers to use multiple channels. “The number of people who have been engaged in more than one source has increased almost six times from 2003 to 2005,” he says.

There also has been exponential growth in the number of shoppers switching from the catalog to the NFLShop.com site, he says. “Our goal is to drive people to the web,” he says.

Not all retailers can afford to set up a sophisticated data integration program like the NFL’s. But there are lower-cost, easier-to-implement solutions available from boutique marketing services providers, predictive analytics and business intelligence vendors, and CRM providers, Ament of Aberdeen Group says.

And retailers with statisticians on staff could set up an integrated database in-house. “The choice is really up to the organization,” she says.

linda@verticalwebmedia.com