If the answer to this question appears to be obvious, please read on! There’s been lots of discussion about how forecasting is done, but much less about what is forecasted. Now, this is an important issue for all forecasters, but none more so than Supply Chain forecasters. Let’s suppose you’re using a particular forecasting method and there’s no scope to change it in the short-term. What can you do to improve forecast accuracy?
As a manufacturer or wholesaler, if you are blindly following your customer’s orders to plan your inventory then you are a victim of the Bullwhip Effect. This is according to the ground breaking research by Hau Lee and his colleagues at Stanford University. The Bullwhip Effect is a demand amplification phenomenon which shows that the more far you are from the actual consumer, the more inventory and/or stock-outs you will experience.
So how do you tackle this? Well, simply by using approaches for being more consumer centric rather than customer centric. The first approach we’ll be considering is Demand Management. For example, by changing order frequencies or batching requirements, demand can be made less volatile and easier to forecast. Even if you make no changes to your forecasting software, accuracy will improve naturally. Strategies can also be introduced to avoid ‘game-playing’ by customers, who deliberately over-order when they know that stocks are scarce. Many of these strategies are applied common-sense, but may yield larger benefits than you would expect.
The second approach is less obvious, has been researched more recently and has even greater potential inventory savings. It is known as Forecast Information Sharing (FIS). This is where the supply chain members either share forecast or produces a single collaborative forecast with their retailers thus giving more attention to the consumer demand rather than the customer demand.
Many companies such as Wal Mart, Procter and Gamble, Hewlett-Packard, Kimberley-Clarke that embarked on supply chain collaboration strategies have reported huge cost savings. Our simulation results on testing the benefits of FIS on sales data of two organisations, a European Grocery Supermarket and a USA computer hardware manufacturer, show similar results.
It is important for organisations to investigate savings in their own supply chain by such simulations. This realisation of the cost savings could provide a more realistic picture for organisations in terms of the expected benefits. This will also help them make decisions on the investments in information technology required for such collaborations. These gains can be estimated by simulation methods discussed in my talk.
The mantra everywhere is the same; huge savings in inventory and stock-out costs are experienced by collaborative initiatives. So when the actions required for being more consumer centric is so obvious what is keeping organisations back. Issues such as confidentiality and trust are huge stumbling blocks that keep companies away from being consumer centric. Experts suggest that although trust builds up after years of partnership, forced compliance by contracts is not an innocuous assumption. Other various reasons for slow embracement of collaboration activities include costs for information systems implementation, misalignment of incentives, forecast volatility, forecast inflation and lack of forecasting expertise. Again there are various suggestions by supply chain experts to surmount these issues.
Join me at IBF’s Supply Chain Forecasting & Planning Conference in London, 1-2 February 2010 to further discuss the above. During the presentation, we will also share the simulation results of testing the benefits of Forecast Information Sharing and will also explore the factors that affect these benefits.
Your comments and feedback are welcome here!
Dr. Mohammad M. Ali
Buckinghamshire New University