How to Think About Big Data in Retail

Big Data is the epitome of a buzzword.  It is overused, often misunderstood, and seems to always have a nebulous air to it.  This is because Big Data is a very broad term.  It can have different applications and meanings for different industries.  One company that utilizes Big Data might do something completely differently from another company that uses Big Data.  Even though it is non-descriptive, it is a popular term, and it seems that we are all relegated to using it until we can find language to better capture the specific situations in which Big Data has an impact.  Hopefully this post will help you to think about how Big Data is impacting retail.

One way to think about Big Data is that it is the ability to track and analyze data (lots of it) in a highly efficient manner.  People have always tracked and analyzed data, but doing so pre-Internet was very labor intensive, and drawing a conclusion from that data required significant guesswork.  Today, with ever present technology and extreme computing power, huge swaths of data can be tracked and analyzed with little effort.  This empowers people to make actionable conclusions that are backed-up with statistically significant evidence.

An anecdote in this New York Times article does a great job highlighting this distinction.  The article explains that New York City's Department of Environmental Protection wanted to find restaurants that were illegally dumping cooking oil into city drains.  Previously the department would have sent inspectors to randomly check on restaurants with the hope of catching an offender in the act.  Instead the city's Office of Policy and Strategic Planning cross-referenced geographic data on drain clogs with data on carting services (that remove the oil legally) to generate a list of potential suspects.  The resulting list was 95 percent accurate.

Certain brick and mortar retailers are now at the initial stages of creating similarly powerful stories by leveraging data.  Previously, collecting detailed information around shopper behavior was nearly impossible.  Paco Underhill, a retail analyst, made a career out of doing so, but his efforts required hours upon hours of watching shoppers on video.  In 1996 Malcolm Gladwell profiled Underhill in The New Yorker and explained that he had "in the past decade and a half...analyzed tens of thousands of hours of shopping videotape."  Back then Big Data was literally Big Data because it had to be stacked in "plastic Tupperware containers practically up to the ceiling."

Today Walmart is doing the same kind of analysis as Paco Underhill on a much grander scale, without the need to watch a single shopper.  The company recently rolled out its "Scan & Go" app to 200 stores (up from about 70).  With "Scan & Go" Wal-Mart can track every item that a customer puts into a shopping cart, as well as the order in which those items are entered (and thus the path of the customer through the store).  Ikea also recently introduced a similar app.  Additionally Nomi, a well funded startup, offers location based shopper analytics by tracking cell phones.          

In spite of its nebulous nature, Big Data is very powerful.  Large retailers will increasingly look to track and analyze in-store shopper behavior because doing so is much more feasible today than ever before and the resulting insights will be invaluable.

No comments:

Post a Comment