You already know grocery and big box stores are collecting data on your shopping habits — what you buy, when, how often, and for how much.
But behind the scenes, Omaha's ConAgra Foods and other consumer product manufacturers that sell to these stores have become even bigger and more sophisticated players in the “big data” game.
They're now borrowing shopper-specific data directly from retailers and crunching it in new ways to better understand consumers and respond to their needs. That's in addition to how they use internal data to forecast demand and increase sales, and use aggregate retail data to plan promotions and strengthen supply chains.
“It's a gold mine of insights and knowledge that nobody has tapped into in a big way,” said Bob Nolan, ConAgra's vice president of customer insights and analytics.
ConAgra, behind brands such as Hunt's, Orville Redenbacher and Marie Callender's, is now studying individual shopper habits at several major U.S. grocery retailers and big box stores. The retailers don't share a shopper's name, demographics or financial information, but do provide all the purchase information associated with a particular customer number.
Now, ConAgra can see how often Shopper 1234 comes into a store, how often she buys a certain product and what products she tends to buy together — pasta and tomato sauce, for example, or seemingly unrelated items, like tomato sauce and diapers. ConAgra can learn how loyal the shopper is to a certain brand, and what makes her switch among brands. And the firm can sort shopper habits by individual store location, seeing differences among neighborhoods.
Now well into its 2014 fiscal year, ConAgra is intently focused on sales growth after a disappointing 46 percent decline in first-quarter profits.
The solution will involve being more competitive on price and investing more in promotions, CEO Gary Rodkin told analysts last week.
Data analytics can help the company target exactly where promotions and discounts are working.
“We need to bend the trends on our market share. It is a market share gain. It's category by category, customer by customer. We've got smarter analytics, and we've got to put them to better use,” Rodkin said.
Consumer goods manufacturers that use data analytics to understand shoppers outperform competitors that don't, according to Progressive Grocer, citing two IBM studies published in July.
“In an increasingly competitive marketplace, the ability to detect subtle shifts that were previously indiscernible is imperative,” found one study, by the IBM Center for Applied Insights and Kantar Retail.
That's because consumers today are not only bargain-conscious, but they're also no longer limited to a few retailers. Consumers, empowered by technology, have more choices — not just the neighborhood grocery, but also the warehouse club, the dollar store, the pharmacy or Amazon.
“The best way of winning their business is not to try managing them: it's to listen to them, understand them and serve them as discrete individuals,” authors of the IBM/Kantar study found. “That requires considerable analytical horsepower, though, and two-thirds of consumer products companies don't have enough.”
To boost its own horsepower, ConAgra made a big investment in its data analytics capabilities starting in early 2012. The firm hired Nolan, a former PepsiCo. executive, for a newly created position. Under his management, ConAgra added a new business function called customer analytics to the same department that houses two existing areas: shopper insights, a group started in 2006 that studies shopper behavior and needs, and category leadership, where the firm works with retailers and other manufacturers to improve the selection and display of various products. The group employs about 125 people.
Under the new customer analytics area, ConAgra has hired a dozen new employees, recruiting from other consumer goods companies and market research firms. Hiring is competitive as retailers and other manufacturers are also stepping up hiring of data analysts.
“These are different skills than we would have hired for in the past,” Nolan said. Some of the employees work in ConAgra offices, while others are deployed directly to retailers' corporate headquarters.
The firm has also leaned on its IT department to expand its in-house data center to handle the additional terabytes of information now coming its way.
With the new data enabling it to drill down to shopper-level habits, ConAgra can put a finer point on work it is already doing to understand the shopper needs and emotions that drive decisions.
Walmart also has amped up its capabilities. Even though the retailer doesn't use a loyalty card program outside of Sam's Club, it has a unique ability to correlate geography with purchase information because it has so many outlets, including Supercenters, Sam's Club, Neighborhood Markets and its online store, CEO Bill Simon said in September at a Goldman Sachs Global Retailing Conference.
“We think that gives us a competitive advantage that others would really struggle to get to,” he said.
A recent Deloitte Consulting analysis of manufacturers' use of big data analytics found that most of these firms lag in developing their analytical maturity, even though their competitive advantage depends on it. The Grocery Manufacturers Association, which sponsored the study, this fall will host its first conference designed to help retailers and manufacturers take a shared approach.
What ConAgra will do with retailers' data depends in part on what its retail customer wants out of the partnership. ConAgra declined to name its data-sharing partners but said the first retailer to share data in February wanted to better understand how people shopped in its frozen food aisles. Frozen food sales are stalled industrywide, and both retailers and manufacturers like ConAgra are eager to see that change.
One early finding was that people who buy one kind of single-serving frozen food tend to buy several kinds of single-serving frozen food. ConAgra could suggest to the retailer that it group smaller portions together, instead of stocking single-serve pizzas by family-size pizzas, for example.
That strategy might work well in one store but not in another, depending on demographics, and any suggestion to change displays would have to be easy for the retailer to execute, Nolan said. “You can't make things more complicated.”
The data may also reveal other shopping “affinities,” for example that people buying pizzas also love buffalo wings, and the retailer might decide to produce a buffalo-flavor pizza, said Christopher Durham, a private brand consultant based in Omaha.
By sharing and studying the data, he said, retailers can use it to inform private brand product development, another area where ConAgra could benefit considering its acquisition this year of private-label manufacturer Ralcorp.
“With retailers, it's not about big data, it's about big answers,” Durham said. “You can have piles and piles of data, but if there's nothing actionable coming out of it, it doesn't matter.”
It may seem strange that a grocery chain or big-box retailer would give up its information, for free, to a supplier. And strange, too, that a supplier would, for free, work with the numbers and offer advice on how the store could improve sales. Historically, retailers were reluctant to share this information, fearing that manufacturers might give the data to a competitor, or use it themselves to enter the market.
But sharing is becoming more widespread as the value of studying the data becomes clear.
Nolan said sharing information takes a sensitive approach. While ConAgra is the one crunching the numbers, the focus has to be on mutual benefit, not just what ConAgra can gain. “It's almost like being a consultant. We want to become the indispensable partner to our customers.”
Then, if ConAgra foods fit into a larger sales plan, he'll talk about how his canned tomatoes or frozen pastas can help the store.
“If they do a better job of managing their frozen department, and get more people down the aisle, ConAgra will get our share of it,” Nolan said.
No comments:
Post a Comment