Inside Customer Management: Analyze This!

The good news is that Gary Andrews, director of manufacturing and engineering at Campbell Soup Company, recently discovered the potential ROI of an integrated business analytics program. The bad news is that it took a multi-million-dollar loss to make the case.

The problem started in 1996 when Campbell moved one set of manufacturing operations out of one plant, making the second plant a sole supplier, says Andrews. At the time, the move -- calculated through a series of manually researched spreadsheets -- made fiscal sense. Very quickly, however, customer demand changed and the plant move no longer made sense. But the cost of running those calculations again was so high and such a hassle that it was not done until late last year. At that time, a new integrated and automated analytics system quickly figured out that the move had cost Campbell millions of dollars a year, building up to an annual loss of $2 million near the end. Campbell reversed the move.

"We now have a lot of flexibility about where we can make certain products and where we can ship them from," Andrews says. "This whole area of a mathematical model of everything in your business is amazing. To have models that are this large and to be able to model 500 products to one thousand customers and 10 manufacturing plants and all the freight lanes connecting them is very profitable."

Companies Pay Attention

Companies are likely going to pay a lot more attention to analytics this year as they find data analysis -- especially in the wake of renewed retail collaboration -- becoming much more complicated and helpful.

Collaboration will play a significant role in analytics because of the wide range of information those partnerships bring in. To an IT manager, that's a treasure chest full of valuable information and customer connections. To those oft-ignored analytics applications, though, that's a world of incompatible and inconsistent data that's just itching to poison result accuracy.

With 2002's technology developments, the consumer goods technology world is going to be a very different place this year. "With the distribution chain, manufacturing, inventory and payment systems, it's not so clear to me that analytics have kept up," says Nobby Akiha, vice president of marketing for software vendor Actuate.

Collaboration raises a hornet's nest of security and proprietary information issues. For example, do distributors get access to excerpts of crunched/analyzed data, all of the analyzed data or even the unadulterated full-blown raw data? "I think it's going to have to start with crunched data, where the benefits may outweigh the risks," Akiha says. "You have to answer the question, 'How far are you going to allow people outside of your organization to have access?'"

Tigris Consulting Principal Jeffrey Ryan also sees collaboration as the most significant factor for analytics purchasing decisions coming this year. The shift towards a collaborative mindset has revolutionized the way consumer goods companies look at their businesses, Ryan says.

The purchasing decision used to simply revolve around the selection of a manufacturer that worked for the consumer goods' own environment. Not anymore. According to Ryan, today's analysis needs to include the environments of many different partners and has to incorporate Internet issues such as collaborative planning, forecasting and replenishment.

Akiha also argues that today's improved ability to grab and extract data courtesy of better customer relationship management software is adding requirements for analytics. "Consumer goods players really do need to be able to monitor trends at the regional store levels better than they are doing today. That would permit much more interesting calls on how they run promotions," he says. According to Akiha, consumer goods IT departments are very good at understanding what might be happening with distribution partners in Chicago and Boston, but can drop the ball when they try and bring the data together to analyze a variety of what-if scenarios. Akiha attributes this to inadequate business analytics.

Multiple Front-Ends

QlikTech CEO Chris Berg says one of the big pushes from the use of analytics will come from its speed and ease-of-use. "Executives now need to have much easier access to data analysis," he says. Business managers can no longer wait for month-end reports to see how products move.

Berg doesn't see that changing the actions of a lot of the larger enterprise consumer goods companies, as much as prompting a lot of the smaller and mid-size players to upgrade systems to more sophisticated analytics. "This will force the move of (high-end) analytics to the midmarket," Berg says, adding that most of the small- and mid-sized players today use whatever happens to be around. "Our experience is that they have been resorting to living with whatever comes with their enterprise application or basic analysis with Excel."

Excel is very popular among consumer goods companies and many analysis firms are taking the path-of-least-resistance approach to placing their analytics engines running transparently behind an Excel front-end. But the distributed nature of today's analytics also means a wide range of front-ends for different kinds of users.

In the Excel arena, the IT people will get the ability to maintain more control over the data, Akiha says. For applications, front-ends could range from a mobile device or Web page to a more sophisticated OLAP (On-Line Analytical Processing) interface.

There are those who argue that another way to make analytics more effective is to make sure as many employees as possible have access to the results and the analytic query capabilities. "For companies to be able to manage all of their business processes, they must deliver analytics to a much wider range of users than they are doing today," Berg says.