Retailing nears holy grail in ‘Big Data’

Name of Publication: 
The Atlanta Business Chronicle
Excerpt of Article: 

By Jennifer Lewis Priestley

The Atlanta Business Chronicle

Friday, Nov. 30, 2012

Yesterday I was online looking for a red, cotton wrap skirt, size 6 (OK, maybe size 8). After viewing several different retail sites, clicking through countless options, I found the perfect skirt. But I had to “abandon” my cart to take care of a minor household crisis. When I went back online, it seemed as if every ad included size 6 women, wearing red wrap skirts. Even more interesting, most came with an incentive for free shipping or 10 percent off.

Clearly this is not a coincidence — does everyone see ads for red wrap skirts? Probably not, but you may see ads for items that I don’t. Welcome to the brave new world of retailing in an era of “Big Data.”

The buzzword “Big Data” is followed by descriptions of terms like “petabytes” — a petabyte is 1 million gigabytes. We have known for a long time that data is cheap and easy to capture and store — inever-larger quantities. But Big Data is about more than just size. It’s about the new kinds of data which are generated from relationships between retailers and consumers using social media, credit card transaction data, click-throughs, and even rate of cart abandonment. And, unlike data from the past, this data is updated continuously — making these types of data more like flowing rivers rather than static lakes.

So, Big Data is about size, yes, but it is also about velocity — updates in real time — and nontraditional, unstructured data. Web data, text data, social media data do not come in as files with clean rows and columns which can be dropped into a spreadsheet.

Some retail organizations are “native” to this kind of Big Data — think Amazon and Netflix. An example which is common to both of these companies is the recommendation engine — “People, who bought this, also bought this ...” Amazon recently reported that over 30 percent of its sales were generated through its recommendation engine. These engines did not exist 10 years ago, because the raw material to power them — Big Data — was not available.

Even “non-native” organizations like Coca-Cola are working through how to redefine their relationships with consumers. Coke just celebrated its 50 millionth customer contact on Facebook. BBD (Before-Big-Data), the idea of a packaged goods company interacting directly with their end consumers — much less 50 million of them — would have been unimaginable. Today, almost all consumer packaged goods companies and retailers interact with their customer base in some form.

Retailing in the world of Big Data is getting closer to the holy grail of consumer marketing — getting the right product to the right customer at the right time at the right price.

The challenge for retailers now is to find the talent — the people who can brave this ever-flowing river or tsunami of data and translate it into a lift in sales — like promoting a red wrap skirt to a price-sensitive size 8 who, in her mind is a size 6.

Priestley is associate professor of statistics at Kennesaw State University and director of KSU’s Center for Statistics and Analytical Services.