Gilt Groupe: What Shoppers WantUsing analytics to bring a personalized touch to online retailing
May 02, 2011
by Sandy Smith
Gilt Groupe sales move quickly, as each event typically lasts less than 36 hours. While that adds to the exclusivity of the online retailer — known for its “insider” pricing on luxury merchandise — it also makes it more vital that the company understands individual customer preferences and what motivates purchases.
“It was an extremely challenging environment for analytics because all of our data assets are stored in multiple locations and various formats,” says Tamara Gruzbarg, senior director of analytics and research for Gilt Groupe. “There was some data stored in-house, some with third-party vendors, and in all formats that we can possibly think about.”
When someone wanted information, engineers had to extract the data and then someone else had to analyze, format and create the reports. “Everyone was looking for data and there was no efficient way to access it,” Gruzbarg says.
Information is extremely important to retailers, but Gilt Groupe’s experience highlights the need for access, especially in the fast-moving online world. Otherwise, the report can be as outdated as that Dolce & Gabbana belted dress — so two days ago.
According to the Aberdeen Group’s December 2010 report, “Data Management for BI: Fueling the Analytical Engine with High-Octane Information,” the typical organization sees a 41 percent increase in data volume year over year. That same report also showed that a company tends to manage about 15 unique data sources.
For a rapidly growing retailer like Gilt Groupe, the data volume can be even more overwhelming. Founded in 2007 as an invitation-only shopping site, Gilt now has more than three million members and has expanded to include menswear, children’s clothing and home décor.
With each expansion, it has become even more important that Gilt “make the browsing experience better for customers,” Gruzbarg says. Because Gilt requires registration before anyone can access the shopping site, the data possibilities were rich; the company just did not have an efficient way to access it. Enter SAS and its business analytics software.
Though the terms “business analytics” and “business intelligence” are often seen as interchangeable, SAS describes analytics as “the ability to predict and optimize,” says Diana McHenry, SAS’s director for global retail product marketing. “Business intelligence is more of a foundation for historical reporting. You layer on the advanced analytics and are able to better understand your customers and their buying behaviors, predict the future and optimize prices, sizes, offers and more.”gilt groupJeansonWall.jpg
Growing cross-category shoppers
Gilt has used analytics to enhance its marketing to customers, making sure they know when an item that may be of interest to them appears. “Not everybody visits the site every day,” Gruzbarg says. “It is important to marry the product to the customer at the right time.
“We have to devise very specific views and aggregations that work around this particular challenge,” she says. “We have to have a robust prediction that allows us to understand the shopper. We may not have that particular brand, but something in a similar style, a similar product and price point that we think you might like.”
With 36-hour sales, there is always the question of what’s working and what isn’t. Having access to the right information has resulted in a significant lift across several categories. Cross-marketing campaigns that target a buyer of apparel who hasn’t yet made a home décor purchase, for example, resulted in a 200 percent increase in multiple-category buyers, Gruzbarg says.
Online retailing certainly offers more access to information, but as Gruzbarg points out, the relatively low barrier of entry to an online store magnifies competition. “This is the only way forward,” she says. “The data is so much richer when you are operating in the online environment. We see sites multiply on a daily basis. Our competition has increased significantly, but we see knowing our customer as a key differentiator.”
Knowing the customer is not just the goal of small- to medium-sized businesses, of course. Nor is it relegated to online retailers.
“These solutions are not out of reach for anybody,” McHenry says. “We really help people crawl, walk and run in terms of business results, analytics capabilities and scalability.”
In many ways, analytics can provide the same sort of personalized experience as when a sales associate knew his customers and would call to let them know of a new shipment or a big sale. “The advantage for us as analytical people is, we can tie all of the site behavior to specific individuals,” Gruzbarg says. “For our entire membership, we understand their visiting pattern, what people like, their first purchase. We also collaborate with third-party providers and acquire a wealth of demographic information on our customers.
“The whole goal of us mining the information is to understand how we can make the experience as personalized as possible for our members,” she says.