What Big Data Means for Retailers
by Jeff Weidauer
Special to The Shelby Report
If 2011 was the year of mobile, 2012 is shaping up to be the year of big data. Big data is defined as datasets too large to be managed using current toolsets or software available to a typical business. Big data will affect everyone no matter what their business, or the role they play.
Big data’s affect on business will be far more than the people in IT. A recent Gartner report predicts that by 2017, CMOs will spend more on data and technology than their CIO colleagues. Big data will be the focus of those expenditures, with the intent of becoming more competitive and in tune with customers.
For retailers, big data will have a significant impact. In today’s world many food and drug retailers focus on price almost exclusively. Customer loyalty is only as good as the last best price offered, and with the typical shopper visiting five or more stores per week, the stakes are high and getting higher. Big data will allow those who make use of it to be more competitive and more profitable by developing deeper relationships with their shoppers, without sacrificing themselves at the altar of low price.
Big data makes information much more transparent and accessible, and at a much higher frequency of input and output. It allows businesses to collect more information with greater accuracy on everything from transaction details to inventories to labor usage.
Most importantly for marketers, big data allows much more narrowly focused shopper segmentation—even down to the individual level—and therefore much more finely tailored communications and offers to those segments.
Big data is a collection of multiple data inputs, from myriad sources. Data sources can be either structured or unstructured, with everything from email messages, social media postings, weather patterns, geophysical anomalies, competitive activity, promotional plans from last year, changes in technology to shoppers and so on all adding to the mix. Bigger is definitely better in this case; as the datasets and the quantities of information they contain multiply, so does the ability to gather useful information that can help accurately predict shopper behavior within a defined set of variables.
In food and drug retail, loyalty programs have long been a staple in the attempt to segment shopper groups by offering insights to the most profitable shoppers and what motivates them. Unfortunately, few loyalty programs actually engender loyalty. Instead, they have become discount programs with little benefit to the shopper or the retailer.
In fact, research done in 2011 paints a bleak picture for loyalty programs. Membership not only doesn’t have its privileges, it’s meaningless for most members. An alarming 85 percent of loyalty cardholders have never heard anything from the retailer since signing up, and 81 percent have no idea what the benefits of the program are supposed to be.
Big data brings with it the perfect opportunity to create true loyalty programs, and provide benefits to shopper and retailer alike. But just as with the original loyalty programs, activating this opportunity requires an investment that far exceeds what most businesses are used to spending.
The original loyalty programs suffered from a wealth of data and a dearth of analytics power. Collecting massive quantities of data is only beneficial if insights are coming from them, and that takes specialized tools and skills.
With big data, that challenge grows exponentially as the datasets expand.
To be sure, the advent of big data will present hurdles, but not participating is not an option for those wanting to stay competitive—and solvent. The use of big data will quickly become a central component of competition and growth. Data-driven strategies will change the competitive field, and for those without the data skills, competing in the new environment will amount to bringing a knife to a gunfight.
The benefits of big data are at least as large as the challenges that must be overcome. Microsegmentation is an obvious and game-changing outcome of big data; the ability to create customer segments down to the individual level will soon be possible.
Even without going that deep, creating segments that number in the hundreds as opposed to the five to eight most retailers have today will offer the ability to connect more fully with the shoppers in those segments.
Gathering insights about shopper behavior in the big data world will include far more than the typical information gleaned by loyalty programs today. Most of that analysis looks at past behavior, compares with like members of a segment and then looks for differences. Or it finds lost categories, i.e., categories where a given shopper used to buy, but now doesn’t.
Assumptions must be made to execute against this level of data, and those aren’t always accurate. Risks arise from assumptions and can cause more trouble than benefit.
The advantage of big data here is the ability to combine multiple data inputs to more accurately understand shopper behavior—and any changes in that behavior—based on data gathered from loyalty cards, social media and mobile usage. This is information that exists today, but putting it together effectively is a complex task, to put it mildly.
Big data also will affect the way stores are merchandised. Some chains today are incorporating store-specific planograms that model the merchandising after the local neighborhood, primarily basing decisions on demographics.
Big data has the ability to add to that knowledge with psychographics, local weather variables, traffic patterns, etc., to create complex econometric models by store that allow dynamic pricing based on external criteria, or cross-merchandising based on what’s on sale this week combined with complementary regularly-priced items.
These are things that can be done today, given enough time and if the right information is available. But gathering the necessary information is a daunting task in itself, never mind actually mining it for opportunities. The toolsets required by big data can take pre-designed business rules and sift through massive stores of information looking for matches based on those rules.
Trying to do this type of collecting, collating, analyzing and finally acting upon the insights generated week after week are just not possible without highly specialized data analytics software, analysts to perform the heavy lifting and marketers who understand how to extract valuable insights and act on them.
The ultimate outcome of big data is a world that might look very much like the one presented in the 2002 film “Minority Report,” in which shoppers are identified uniquely and presented with specific, relevant offers based on shopping history and other data elements. In fact, the scenario presented in that movie is less than what is possible using big data, although the retinal-scan technology is less likely to be a part of a supermarket loyalty program.
The promise of big data lies in the ability to market more effectively with less waste and with a much greater ability to predict results beforehand. The challenge is in finding the necessary tools and skills to deliver on that promise.
The one sure thing is that big data promises to change our world—indeed, is changing it already. And much like social media, whether one decides to participate in it or not, the conversation is going on now.
Jeff Weidauer is VP of marketing and strategy for Vestcom International Inc., a Little Rock, Ark.-based provider of integrated shopper marketing solutions. For more information please visit www.vestcom.com.