Retailers Help Del Monte See the Future

March 22, 2010
CIO Marc Brown shares strategies for improving demand-planning accuracy.

The recession has resulted in some tough lessons for manufacturers about demand planning and the uncertainty inherent with forecasting models. A recent AMR Research report suggests the economic downturn should have served as a "wake-up call" for the importance of sales and operations planning (S&OP).

An S&OP plan is typically characterized as a consensus process that involves stakeholders internally and externally throughout an organization, says Jane Barrett, research director at AMR and co-author of a January report entitled "Conquering the Seven Deadly Challenges of Sales and Operations Planning."

But few companies seem to be getting the message that S&OP can help lessen the impact of demand fluctuations, according to the AMR report based on 182 interviews with U.S. firms in August 2009. Two-thirds of the companies surveyed in the AMR report say improvements to their S&OP processes have stalled, and only 18% of the firms say they have an extremely effective S&OP plan.

One company Barrett cites as a model for a successful S&OP plan is processed foods manufacturer Del Monte Foods Co. The $3.6 billion company made changes to its demand-planning process in 2006, leading to improved visibility and forecasting accuracy.

IndustryWeek spoke with Del Monte Chief Information Officer Marc Brown for an article on forecasting that will appear in the May issue of IW. Here are some excerpts from IW's conversation with Brown:

Marc Brown, Chief Information Officer, Del Monte

IW: Take me through some of the ways you've addressed some of your forecasting challenges?

MB: We did a couple things. Number one is the ability to use downstream data as an additional demand signal. So we were already getting downstream data from a couple of customers, but we weren't using it to the fullest extent in our demand planning or, for that matter, in our deployment planning and some of the other activities that we do in S&OP. So we started to use that data. We put in place the repository for it and built executional tools to be able to use that data as part of our enterprise volume forecasting process, so not only were we doing a demand-planning statistical forecast as well as a sales-informed sales forecast and a marketing consumption-based forecast, but we were also doing a retail store-level forecast. Initially, looking at retailer distribution centers and folding back into the store level, and then we have the ability to back propagate that demand and see it at our geographic distribution points.

IW: How are you accessing that data and are you analyzing it?

MB: We pull the data in from the retailers data source. So each retailer has their own. Wal-Mart has a tool called Retail Link, which provides sales, inventory and forecast information at store level, and several of our other retailers provide similar data using whatever their approach is for providing that data. We bring the data in. We normalize it so that we can use a common repository, and then we use a tool called One Network where we back propagate that demand from the retailer supply network into our supply network, so that we can project where is that demand going to hit in our network, and then within that One Network tool we also have visibility into our inventory to our order volumes to all of our internal data, so in one place we can see the entire supply chain -- both what's actually happening, how much inventory is in stock at a given time, as well as what we're projecting to happen in the future. So one component of that is to feed that forecast visibility into our enterprise volume forecasting process, which we run in a product called i2. In i2 we bring in that retailer forecast as well as our marketing consumption forecasting that our marketing people are doing. That gives us the ability to have a discussion across the business and determine what our view of the future is going to look like. Now, within that also we can measure forecast accuracy and bias and all those kinds of things so we've gone from being pretty poor at that forecasting to being really very good at that forecasting.

IW: How were you forecasting before?

MB: We were using some statistical modeling and limited view of our trade spend plans. So it was far less robust.

IW: Have you noticed significant improvement in terms of forecast accuracy?

MB: Some categories we've had better success than others. We're in the high 80% accuracy level across the majority of our brands. We were probably in the 50% to low 60s on average before we started this. From a service-level standpoint we're in the high 99s. That means we shipped what the customer ordered, and they got it in time, which is fabulous. It's best in class. So what it says is by improving the forecast and improving our deployment capability as well as our supply planning capability using that data in our forecast, we've been able to reduce inventory, raise service and we're seeing it in our forecast statistics.

IW: Were there challenges getting retailers on board?

MB: It varies by retailer. Some not only make it available they want you to have it and they want you to use it. So they're pushing. Some retailers want to do it but aren't capable yet, and some retailers don't see the value. Now, one of the things that we've done that we've started to do because we believe it's important and an important role that we play in the supply chain and in the food-distribution channel is we've started to work with retailers on how to improve the flow in their network and how to improve their cash flow against our products. So because we have visibility to this data, we can help them to understand where they can improve the cycle time of our replenishment and as well as the in stock level at shelf and get some of the same type of inventory reductions that we've been seeing with equal or better on-shelf availability, which is pretty powerful.

IW: How do you handle the retailers that aren't capable or don't see the value?

MB: To some extent what we've done is built our capability to use as much data as we can get, but if we don't get any data we can still execute. So we're pretty busy with the retailers that do see the light and want to make this stuff happen, and we try to communicate with the ones that don't and help them to understand the opportunity and help them to get whatever they can get out of our capabilities. But you can lead a horse to water, but you can't make them drink. So we're not dependent on it but given that we have it we believe we can be best in class and that they can get better profitability with our products, which is usually a compelling message. And in this economy cash flow is a huge objective for our retail partners. So there's the opportunity for them to sell the product to their customer before they're paying us for it or at the same time they're paying us for it. That's pretty powerful. And we're doing that with several retailers.

Upcoming: In May IndustryWeek will feature several manufacturers, including Del Monte Foods and Dow Chemical, in an article exploring current forecasting trends throughout manufacturing.

About the Author

Jonathan Katz | Former Managing Editor

Former Managing Editor Jon Katz covered leadership and strategy, tackling subjects such as lean manufacturing leadership, strategy development and deployment, corporate culture, corporate social responsibility, and growth strategies. As well, he provided news and analysis of successful companies in the chemical and energy industries, including oil and gas, renewable and alternative.

Jon worked as an intern for IndustryWeek before serving as a reporter for The Morning Journal and then as an associate editor for Penton Media’s Supply Chain Technology News.

Jon received his bachelor’s degree in Journalism from Kent State University and is a die-hard Cleveland sports fan.

Sponsored Recommendations

Voice your opinion!

To join the conversation, and become an exclusive member of IndustryWeek, create an account today!