Can Smarter Computers Give New Meaning to Marketing Intelligence?
By Alison Otto, Prepared Foods
Suffering from information overload?
You’re not alone. Since the arrival of scanner data in the early 80s, food companies have been awash in information. In fact, A.C. Nielsen estimates, scanners have increased the amount of data food companies receive by 500 to 32,000 times.
Relief may be in sight. Chicago-based Information Resources Inc. (IRI) and cross-town rival A.C. Nielsen have both developed new expert systems to help food companies glean more meaning from scanner data. It’s the beginning, they say, of systems that will help marketers and sales people formulate strategies quicker and with better results.
Says Gary Overhultz, an IRI executive vice president: “The use of information for competitive advantage is just blossoming.”
The systems, such as IRI’s CoverStory and Nielsen’s Advisor series, are designed to eliminate the drudgery of sifting through reports. They zip through scanner data and give marketers answers to basic questions: How’s business doing? Why is it doing what it’s doing?
IRI’s CoverStory, for example, pulls out the “news” in scanner data and spits it out in a one- or two-page memo. It may point out category volumes and shares, or present more focused analyses of shifts in brands. The system doesn’t recommend strategy, but it does alert marketers to where sales are up or down and what factors accompanied the changes.
Nielsen’s Marketing Navigator, due out later this year, will give marketers what company vice president of product planning and development Danny Moore describes as a “guided tour through data.”
For sales people, IRI’s SalesPartner is designed to sift through data and suggest key selling opportunities. If a company’s business is down in Ft. Worth, Texas, for example, the system might suggest that one chain is carrying only five items while another is carrying nine. Then sales people can gear selling strategies toward improving volume in that particular store. Nothing but DRIP.7
Makes sense, doesn’t it? But there are some kinks.
Syndicated data services have moved data retrieval light-years beyond where it was five years ago. While industry experts generally are impressed with the new systems, they suggest that food companies still are struggling to make rhyme or reason of the piles of data they have.
“They’re DRIP-data rich and information poor,” says Mike Moriarty, a principal with A.T. Kearney, a Chicago-based management consulting firm.
“On a scale of 1 to 10 on their ability to use scanner data to its maximum, I don’t think any company is better than a 3,” adds Mick Siegel, another principal with A.T. Kearney.
Steve Rubinow, Quaker Oats’ director of decision support services, sees Quaker as maybe a 5 on that scale. However, he says, the trick is to move from data to information to insights.
“Even when it’s information, it’s still too voluminous,” he adds.
What’s more, marketers say, although companies have made progress with these new systems, it will take time before they figure out how to use them for their real purpose: gaining competitive advantage. Keeping it simple
Some companies have tackled the problem with a narrowly focused approach. Ocean Spray, for instance, has been using an expert system called Distribution Agent since the summer to help pinpoint sales opportunities. The system feeds scanner data into a sophisticated computer model developed at Duke University.
DistributionAgent sorts through scanner data, market by market, looking for gaps between actual and “deserved” levels. When it finds a gap, or an opportunity, it formulates a selling argument.
The results are distributed to the sales force without them ever having to sit at a computer. “We joke that our user interface is an envelope,” says Gordon Armstrong, the company’s manager of syndicated services.
Still, warning of “infocarcinoma,” Armstrong says he purposely tries to keep it simple. DistributionAgent can’t handle unique questions, an approach he advocates for now. “Our goal is automation, not flexibility.”
Pillsbury also is taking a new approach to sales. The company’s frozen food sales department automated about 15 months ago with one goal in mind: category management. “We’re creating systems to align ourselves with our customers,” says Bud Floyd, Green Giant’s director of sales planning.
With the help of a consultant, Pillsbury developed its own system to give sales people the kind of information they need to sell frozen, foods to retailers in a way that helps improve retailer sales and profits-and move more Pillsbury products at the same time.
While the program is still in its infancy with a few key accounts, Floyd reports some encouraging results. In one case, he says, the company analyzed a customer’s distribution and promotional history and the way it buys products. The presentation was persuasive enough for the chain to run a big promotion that moved half a year’s worth of product in only 10 days, Floyd says.
Pillsbury’s sales people are being trained on computers. It’s a slow process that requires sales people to think more analytically. Floyd advocates starting out analyzing only a few variables. “Think of the system as a tool kit,” he says. “You don’t always use every tool in the kit.”
Although some companies are chipping away at portions of the information puzzle, some experts suggest there are bigger issues to be resolved before expert systems can work.
What is holding up progress?
There are a couple of flies in the ointment.
Defining the rules. Technology is spinning out a lot of answers before companies know exactly what questions they need answered, says Jim Morehouse, an A.T. Kearney vice president. “In the absence of a target, everything is muddled,” he says.
Eric Deaton, manager of research systems for Kraft USA, suggests that companies have to decide which facets of the business need to be run on information systems and which ones don’t.
“There is still a lot of wisdom and art in using the marketplace information,” says Deaton. “Until we figure out the uses (for scanner data), there’s no sense in refining the mechanics of it.”
Adds Nielsen’s Moore: “There’s incredible interest in the whiz bang and what it might do.” However, he acknowledges, “Every company is in kind of a unique operating environment.” v Training. It’s crucial to put the information into the hands of those who need it. But food companies are reticent to make computer jockeys of their sales reps and marketers. “We still expect the sales reps to think like machines,” says Deaton. “We need the machines to think like sales people.”
Quaker’s Rubinow says food companies may have to recognize that learning and using the systems will take time on the part of those who use them. The alternative is to have data crunchers who would do the groundwork. He cautions, however, that predigested information might be too simplistic-or too easily misinterpreted.
For the new glitzy system to work information companies first have to sell food companies on the idea of adapting a new computer mindset.
It can be a tough sell, says IRI’s Overhultz, especially to overcome doubts within what he calls the jaded user community,” the people who have listened to the promises of decision support systems over the last 10 years that didn’t deliver.
Meanwhile, it’s a confusing time for food companies. While many companies are experimenting with expert systems, most say they haven’t yet hit on just the right combination of technology and street smarts.
The new systems are smart, but maybe they’re not smart enough yet. Says Ocean Spray’s Armstrong: “It’s fairly easy to generate information, but hard to generate insight. We need more meaning.”