Before the advent of Big Data, mass -advertising was comparable to the pitches of those fragrance sellers who stand at the entrances of department stores, singing out lines like: Are you interested in a floral scent? How about a woodsier one? I have a great new citrus fragrance I think you’d really like!
“For many, the immediate response to this firing of questions is to run a million miles in the opposite direction — especially if you’re male,” said Christopher Sanderson, co-founder of the Futures Laboratory, a London-based trend forecasting consulting firm.
Yet, for many years this was the sort of advertising model that retailers operated under: Blast consumers with large-scale promotions and big, 100-page mailers and hope something will catch the eye. Increasingly, however, retailers are turning away from relying on hope to grab attention, thanks to Big Data — a term that refers to the trail of information we leave each time we make a purchase online, “like” a retailer’s Facebook post or allow our phone to track where we go. Using in-house statisticians or outside firms, retailers are getting a more nuanced view of what their customers like, where they live and the sorts of things they buy.
“With Big Data, retailers can create a 360-degree view of who their customers are,” said Mark van Rijmenam, a Netherlands-based strategist. “They can then create the right products for the right price for their customer.” Data analytics — the science of breaking down and interpreting all that “big data” — can pinpoint where people like to shop, how much time they spend in a store and which streets they frequent most often. Such information helps developers and brokers determine, for instance, the best location for a new store, or where to place a billboard for the maximum number of views.
Could Big Data put an end to those aforementioned perfume-advertising gauntlets? In a 2015 report about the £345 billion (about $520 billion) global beauty market, one of the -largest trends Futures Laboratory cites is the increasing use of Big Data on the part of beauty brands to further personalize, target and communicate with consumers. The report notes beauty company ModiFace’s new app, called Beautiful Me — which helps users select makeup, perfectly matched to their skin tone, through scanning some 500 Facebook photos — as an early example of the use of Big Data in cosmetic selection.
To show how the phenomenon can also help influence brick-and-mortar decisions, the report describes the way e-commerce site Birchbox used customer predilection to design its first physical store: a 4,500-square-foot foot duplex in New York City. “We applied the insight and feedback from our hundreds of thousands of customers to better understand what makes them tick and create a customer-first, holistic offline shopping experience,” said Birchbox co-founder Katia Beauchamp.
What Futures Laboratory concluded, overall, is that random, nonanalytical targeting of consumers no longer works. “Over the rest of this decade, the ability to read the runes in a torrent of data will separate the beauty winners from the brand also-rans,” the firm said. Van Rijmenam was more blunt. “Retailers, in order to survive, will have to start adapting to this,” he said.
The idea of using Big Data to better target consumers has excited retailers and companies since the 1990s. Many large companies have long kept tabs on customers — tracking, for instance, any time someone has phoned the customer-service line or redeemed a coupon. The change today involves the amounts and types of data that retailers have access to.
“The data sets from which we are trying to learn useful things has just exploded in terms of volume and diversity,” said Vasant Dhar, co-director of the Center for Business Analytics at New York University’s Stern School of Business and editor in chief of the journal Big Data. Indeed, the amount of digital data being produced globally is doubling every two years, according to International Data Corp., a Framingham, Mass.–based consulting firm. Meanwhile, the tools, mechanisms and analytics to decipher these reams of data have become more advanced, with large retailers banging down the doors of university math departments. “It’s like an arms race to hire statisticians nowadays,” former Amazon.com chief scientist Andreas Weigend told The New York Times.
Walmart was among the first retailers to start collecting reams of customer data. “Walmart was using big data even before the term Big Data became known in the industry,” said van Rij
menam. In 2011 WalmartLabs, the company’s innovation lab and research-and-development center, launched Social Genome, a tool to help suggest purchases to customers. Social Genome arrives at these suggestions by combing through customers’ social-media posts and collating that information with data it already has in-house. If a customer in New York City tweets that he is hungry and thirsty, the company uses a fast-processing system it calls Muppet to sift through its data collection and discover that this user may love pizza and beer — whereupon the retailer emails him coupons redeemable at a nearby store.
Amazon is another famous user of Big Data. Whenever an Amazon customer looks online at a specific item, the company’s computer platform automatically makes suggestions for similar items of possible interest, based on buying patterns identified among other customers. Such recommendations have accounted for as much as 20 percent of Amazon’s total sales, says Dhar.
But this much discovery and revelation of personal information can cause trouble and may backfire. In 2012 New York Times reporter Charles Duhigg revealed that Target had created an algorithm which could, based on a woman’s spending patterns, predict with a high degree of accuracy whether that customer was pregnant, and when she was due. Target sent related coupons and ads to all these customers — one of whom turned out to be a teenager in Minnesota whose family had not known she was pregnant. The story went viral, and privacy advocates, alarmed over the ways such information is being collected and used, made a target of Target. In January a law went into effect in California prohibiting online companies from compiling or using personal information about minors, and there are about 80 similar bills being considered this year across 32 states, according to the Data Quality Campaign.
But futurists argue that people are yearning for more personalized, targeted marketing, not less. Only people over age 50 care about protecting their privacy, the futurist Sanderson told ICSC’s European Conference in April. “I’m a vegetarian,” he said. “In 10 years I’ve never bought a single meat item. Yet [the grocery store I shop in] still hasn’t gotten the fact that I am not likely to be tempted by ham or sausage or salami, and is always sending me coupons for these things. It’s annoying and makes me think they don’t care to learn anything about me.”
Today most large-scale retail companies are investing in Big Data collection and analytics in some way. But experts say it is impossible to know exactly how many retailers are using this, the ways they are using it, and whether or not such use has been successful. Companies are being secretive, for fear of alarming customers or tipping off competitors. (Several declined interview requests for this article.) As a result, there is no quantified system or benchmark to measure success. “I see a lot of interest in using Big Data to generate value, but I see very little in quantified data” related to levels of success, Dhar said.
This lack of information can foster anxiety in retail owners. “The landscape out there is one of hope and expectation and worry,” Dhar said. “There’s the hope that companies will be able to derive a lot of value from the information, but there’s also the worry that they’re not moving fast enough and that someone will disrupt them out of existence.”
Right now, most companies using Big Data are drawing conclusions about consumers’ buying patterns and behavior by studying and analyzing mass data trends, rather than tracking people individually. The analytics are not quite fully in place yet to help determine whether the person seated at a computer in New York City and who ordered a J.Crew dress last week is the exact same person who is just now, two weeks later, entering the Michael Kors store at The Grove, in Los Angeles. “The holy grail is to stitch together a customer-centric view, which is hard to do based on the fragmented information that flows in from various channels,” said Dhar. “There is, though, all kinds of information you can buy from third-party providers to integrate with information you have.”
AirSage, an Atlanta-based provider of population and location analytics, is among the nation’s largest sellers and aggregators of Big Data related to travel information. AirSage works with two of America’s top-three cellphone carriers to put software that is embedded inside the data center’s of the carriers. At any given moment, AirSage can tell where a third of the U.S. population happens to be — though it strips out all the personal information from the data.
For the past 18 months, AirSage has sold these data sets to reatilers, brokers and real estate developers to help with site analysis. “We can tell you, at any point, how many people might be traveling past a certain location — and how important they are to your target market,” said Andrea Moe, AirSage’s vice president of product management and marketing. “If a restaurant chain that’s located right off the highway, for instance, wants to put up a billboard, we can tell you how many residents in the area pass by it, versus how many visitors.” Though AirSage declines to name any businesses it has contracted with, Poe says the company helped one Los Angeles mall chain better understand the demographics and activities of shoppers at 10 competing shopping malls. The chain used that info to change its marketing strategy, she says.
MasterCard Advisors, the professional-services arm of MasterCard, entered the commercial real estate arena in May with the launch of Retail Location Insights. Analyzing and collating anonymous data gleaned from card usage — the company processes about 160 million transactions per hour at some 38 million merchants worldwide — Retail Location products provide a look into aggregated sales at retail locations across the U.S. and Canada.
“The retail commercial real estate market has relied for a long time on third-party, self-reported and anecdotal sales data to measure the revenue performance of locations, so it’s been a very nontransparent market,” said Gary Kearns, MasterCard Advisors’ group executive for information service. “We saw there was a big void we could fill by providing transparent and upfront scores and insight into something as small as a census block.” Retail Location Insights provides scores on five points, including a site’s stability, performance and traffic. “There are consumer credit scores that help banks determine credit risks to lend money,” said Kearns. “We want to provide that same kind of transparency and service for the commercial real estate market.”
Analysts believe that Big Data has the potential to transform the way retailers operate, but only once consumers become more comfortable with the idea. Sanderson likens the concept to the introduction of the cellphone. “Early on most people were very dismissive of mobile phones,” he said. “They saw them as intrusive to their business and personal lives. They didn’t want to be accessible every minute of time.” But the thinking changed as people began to realize the freedom and benefits cellphones were bringing, he says.
It is fear and a lack of transparency that causes consumers to shy away from providing information on their preferences. “Retailers need to prove to customers that, when they click a box, the information they’re providing will provide a more engaging, beneficial experience for them,” Sanderson said. This means less spam, coupons for items a customer actually wants to use, and a more open environment that allows customers to talk to retailers, and retailers to talk back to customers.
“Most people will buy into it,” said van Rijmenam. And as Big Data becomes more universal and accurate, forecasters predict that retailers will have the ability to create individualized portraits of consumers, to know when a customer is about to enter the store, and to have correspondingly suggested items prepared for them when they do enter, and at a discounted price — all of which are the things consumers value, says van Rijmenam. “There will always be a group of customers that don’t want their data used,” he observed, “but they will be the ones who have to pay more for everything.”
