At Continental Foods, the demand forecasts by Supply Chain and Sales kept drifting ever-further apart. Something had to be done. Charlotte Corneillie, the company’s demand planning manager at that time, took up the gauntlet. How did she obtain the right input from Sales, how did she keep a tight rein on the process, and how did she manage to save time while also improving the forecasting accuracy?
Help, the forecasts are drifting apart
“For many years, Supply Chain and Sales issued very similar demand forecasts. Finance consolidated the numbers where necessary and used the input for the balance sheets. But about eight years ago the forecasts started to drift further and further apart. In Supply Chain, we relied on the statistical forecast which is based on historical data. But the market was changing, there was talk of a recession, and that made Sales less optimistic. The gap between the two versions became so big that something had to be done. We couldn’t all continue issuing forecasts, each from our own ‘island’.”
Taking control
“Because it’s my responsibility to provide a realistic forecast for our production facilities, I was asked to switch to a collaborative demand forecast. I could use the input from Sales to enrich our statistical forecast. That presented me with a number of challenges. How could I maintain control? I didn’t want Sales to be dictating all the numbers. And how could I get buy-in from Sales? After all, no one wants to have to do a lot of extra work and they wanted to be certain that their customers would receive optimal service.”
Involving Sales
“Collaboration revolves around being able to put yourself in the other person’s shoes, so I thought carefully about which aspects were important for Sales. How could I get precisely the information I needed without taking up more of their time? How could I make my Sales colleagues see why that particular information was so relevant, to reassure them that we were taking the right factors into account? And I also knew how important it was for them to be sure that their key accounts would continue to receive good service, so I paid attention to that too. During the session on 28 June I will explain how I did it, and how it has resulted in a significantly better collaborative forecast. Last but not least, I’ll reveal an initiative that began spontaneously but ultimately played a key role in gaining acceptance: our Duvel contest. That’s right, the beer brand.”
Result
“The forecasting process has become easier for Sales because I’ve increased the segmentation. They provide very targeted input and can see how it is used. This creates trust and results in a demonstrably better forecast. Furthermore, the structured collaborative meetings enable us to identify and analyse bias; if our last estimate was incorrect, what was the cause of that and how can we avoid repeating that mistake? We now achieve much better customer service levels for our limited editions. I can show our production facilities that we have thought carefully about our forecast, and in the case of capacity restrictions it helps us to set the right priorities quickly. And as the icing on the cake, we now have proactive dialogue. If Sales notices a change in the market, they call me so that we can challenge the forecast together.”
From Excel to effective
“This methodical approach was important, but we still had a bottleneck: the systems. We had a stand-alone system with a single overall statistical forecast for all products. To prepare our demand forecast, we extracted information from that system, put it all in Excel and then set to work with it – from one department to the next. It was even more complex at international level because different countries worked with different systems, so the Excel files had to be altered again for input into their ERP systems. There was a high risk of human error and it entailed a huge amount of work. Nowadays we use uniform tooling which is the same across all countries, so we all speak the same ‘language’.”
