Pulling More Meaning from Big Data


Safeway's digital advertising messages for loyalty club members have been so effective, the retailer plans to eliminate the printed circular it has relied on for decades.

"As people become more digital, there's an opportunity, which we're working hard at, to actually get out of the paper ads and make the ad itself personalized for every household," Chief Executive Officer Steven Burd told investors during his fourth quarter 2012 earnings conference call in February 2013.

Burd said Safeway's Just for U loyalty program had 5.4 million households signed up by the end of 2012, accounting for 45 percent of its sales base. "What's happening is, our best customers are becoming increasingly more loyal and buying more items per trip," he said. "And that's been true…since the beginning of Just for U."

Safeway's decision to tap its shopper data to create personalized ads is not exceptional – the technology to do so is widely available – but it is rare. In fact, experts say, most retailers don't take advantage of the bulk of data available to them to develop insights about their customers.

"Some retailers are more progressive than others, but for the most part, the industry is not prepared nor totally aware of the amount of data or strategic implications of the 'new normal' when it comes to the competitive advantage of consumer and shopper insights," says Marcus Shingles, business insights and analytics practice leader for consumer products at New York-based consulting firm Deloitte.

The data is flowing from a growing variety of sources, but how to convert it to meaningful information is more complex. "Beyond a doubt, the premise that the industry has just begun to tap the data to which it has had access, let alone emerging sources, is correct," says Dave Shuman, co-founder and vice president of product strategy at Orchestro, a big data consulting firm based in McLean, Va. "The grocery industry is drowning in data, and the spigot has just barely been turned on."

"The grocery industry is drowning in data, and the spigot has just barely been turned on."

– Dave Shuman,


Where is the Data Coming From?

Data is generated at nearly every step of the grocery supply chain, from the factory to the store to the Facebook message the consumer posts about how much she enjoyed the fish sticks mom made for dinner. All of this data – in the form of syndicated data, POS data, loyalty club data, data from smartphone apps such as grocery lists, and social media chatter– is available to savvy CPGs and retailers.

"A.G. Lafley [Procter & Gamble's CEO] spoke of the two moments of truth," says John Ross, president of Inmar Analytics based in Winston-Salem, N.C. "The first occurs when a consumer buys a product, and the second when they use it. Much of the data today is about orchestrating and understanding those two moments. But two additional moments of truth are emerging to bookend Lafley's. One occurs when a consumer is planning to make a purchase. The other happens following use, when the consumer talks about his or her experience with the product. All of these activities leave a 'data wake' that describes how the consumer is moving down the path to purchase."

Like most consumer packaged goods companies, Procter & Gamble relies on data to determine what consumers are looking for. "Consumer insight is at the core of our business model. We approach every brand we make by asking the question, 'What do people really need and want from this product? What does this mean to their lives?' Let me be clear – this is not casual observation. We employ teams of behavioral scientists, researchers, psychologists, even anthropologists to uncover true insight based on intensive research and exploration," said Marc Pritchard, P&G's global marketing and brand building officer, speaking at the Association of National Advertisers' 2012 Annual Conference.

Too Much Information?

Most firms haven't advanced beyond localized analytics and don't fully capitalize on the existing data they have at hand – such as POS data, loyalty club data and social media traffic – according to a 2012 Deloitte study for the Grocery Manufacturers Association. "You wonder if they'll have the bandwidth to take on this whole new array, volume and variety of data that is coming," said consultant Tom Davenport in a post-study presentation of the data.

The rapid pace of big data innovation may leave some companies in the dust, while others that can successfully tap the data will succeed. "As we talked to a lot of companies, I think there was a little bit of skepticism about it," Davenport said. The expansion of digital technology in virtually every facet of life is creating an unprecedented amount of data. "There's going to be a massive amount of unstructured data coming at these organizations, and if they're not ready, other companies are going to be able to make a lot more hay with those capabilities."

The companies most likely to succeed with big data are those willing to share it with partners, Davenport said. The cultural change must start with leaders who believe in the power of data analytics and are willing to experiment with the scores of big data innovations that firms can choose from.

One company considered highly successful is Kroger, which has maintained a partnership with London-based dunnhumby Ltd. for more than a decade. The U.K.-based firm helps Kroger mine its shopper data for intelligent insights, which it puts to use in personalized mailings to customers and other targeted promotions.

In 2012, Kroger mailed to 11 million households, providing recipients with "an individualized set of coupons for the products they like and buy regularly – almost no two were alike," the company said in its annual report. The Loyal Customer Mailing achieved redemption rates topping 60 percent, the report says.

Key Applications

Deloitte has identified 80 evolving innovations in big data affecting the industry, Shingles says. Underlying most is the ability to better understand customers' characteristics and behavior.

"Folks like Sobeys, a leading Canadian grocer, are innovative in their use of firsthand customer data to drive better customer satisfaction, operations, loyalty and results."

– Diana McHenry,

SAS Institute Inc.

"Folks like Sobeys, a leading Canadian grocer, are innovative in their use of firsthand customer data to drive better customer satisfaction, operations, loyalty and results," says Diana McHenry, director of global retail product marketing for software provider SAS Institute Inc., headquartered in Cary, N.C.Sobeys gathers data from its Club Sobeys loyalty program, which allows shoppers to earn points that can be redeemed for money-off purchases in the store, items from a catalog, air miles and other perks.

"New capabilities, like data visualization, high performance analytics, plummeting hardware costs and increased computing speeds mean grocers can be empowered to engage customers and better understand their preferences," McHenry says. "This results in better retention and revenues."

Customer data allows retailers to create personalized marketing material, as Safeway does with its Just for U coupons and specials promoted online. To achieve personalization, retailers must know who their best customers are, what they buy and how to influence them.

"If I am able to understand who my most profitable segments are, that information can help me launch a campaign."

– Bill Tancer,

Experian Marketing Services

"If I am able to understand who my most profitable segments are, that information can help me launch a campaign," says Bill Tancer, general manager of global research for San Mateo, Calif.-based Experian Marketing Services.

Experian helps retailers classify customers into categories of "mosaic types," largely based on where they live, so that companies can create more targeted marketing campaigns, Tancer explains. Safeway has used Experian's analytics to determine that the leading customer type using its website is the "Age of Aquarius" type. This category, which represents 2.3 percent of all U.S. households, consists largely of health-conscious empty nesters with an income of $150,000 to $175,000 who enjoy the opera and belong to art associations.

The more information available about a customer, the more targeted a retailer's marketing message can be. For example, if a retailer knows a shopper prefers a certain brand of coffee, it might feed a mobile coupon to her smartphone for a competing brand just as she enters the coffee aisle. Identifying a customer by her historical purchase patterns and location creates a powerful marketing combination.

But is the data collection getting out of hand? Several New Albertsons Inc. banners acquired from Supervalu by Cerberus Capital Management recently halted their loyalty card programs. Jewel-Osco President William Emmons explained the retailer could do as well simply by examining storewide data.

Other retailers are trying innovative systems that avoid collecting the personal data some consumers are sensitive about protecting. Kirkland, Wash.-based Visible Brands is testing a new Wi-Fi and radio frequency identification (RFID) system in a Yoke's Fresh Market in Spokane Valley, Wash. The system anonymously tracks the progress of shopping carts through the store. Small touch screens on store shelves deliver coupons that can be tailored to the shopper near it based on the shopping cart's path through the store. When a customer touches the screen and selects a promotion, the RFID system automatically records it on that shopper's cart, and the coupon is electronically redeemed at checkout.

Following a customer's path can help marketers entice customers to buy products that complement those already in their basket. "For example, we can know that when buyers are purchasing one thing, say spaghetti sauce, they also buy pasta, capers, bread or wine," Ross says. "New technology enables us to know these patterns, identify the purchase as it is happening, and offer something to complement the purchase, in this case, bread or garlic. Consumers make decisions in real time, [and] technology must keep pace."

Retailers also can use consumer insights to make decisions about which products to stock and where to place those products on the shelf. POS data can be combined with information about upcoming promotions, new products, advertising campaigns, etc., to create the most effective stocking strategies.

A key to using the plethora of data is the ability to react quickly. Consumers make decisions in real time, Ross says, and big data applications are more effective if they can react quickly. Retailers have been using insights to drive their business for years. The traditional approach has involved looking at data in hindsight, he says. "We can now look to data to be predictive as the shopper is planning and making decisions," he says. "And, it can be done in real time at the speed the shopper is moving."

Ed Avis has written for Crain's Chicago Business, the Chicago Tribune, Specialty Coffee Retailer, Tea Magazine and many other consumer and trade publications.

This ad will auto-close in 10 seconds