Project-based business has a high impact on a factory’s demand planning. Typically, there’s a high degree of uncertainty about the effective implementation of the projects, completion is often delayed and, once projects have been won, large volumes of products must be supplied at the same time which results in high production peaks.
In the FMCG sector, supply chain managers and sales managers are used to dealing with promotional campaigns – which can also cause a sudden peak in demand – but in B2B I rarely see a disaggregated forecast showing the different kinds of demand. Highly project-driven businesses typically use project-based forecasting methods, and companies with a significant make-to-stock process focus on long-run rate forecasts. A mixture of both types of forecasts is still a rarity. I wonder why, because a split forecast has a lot of advantages.
Good demand planning starts with making a distinction between project demand and recurring demand, because that’s the only way to improve the forecastability and to enlarge the trusted baseline forecast. In doing so, it is important to clearly define what is a project and what is not by agreeing on the normal order amounts. If a normal customer order consists of 10 pieces then an order of 100 pieces will obviously disrupt your production and therefore must be regarded as a project order. Treating the project orders separately will give you greater demand stability and a higher forecast accuracy for the recurrent orders.
Once a separate project demand forecast has been made, supply chain managers and sales managers must agree on the triggers for inventory decisions. For example, they may decide that Supply Chain should start keeping stock for prospective deals with a closing probability of 75% or higher.
But what’s in it for Sales? If Supply Chain statistically forecasts all the recurrent orders, Sales just has to focus on the project-based exceptions. This will save Sales a lot of time which can be spent on more valuable sales activities.
Furthermore, try to convince your colleagues in Supply Chain and Sales not to hold stock for project orders, and certainly not for completely customized products. This will have a positive impact on your working capital, your inventories and your obsolete stock alike.
Sometimes, of course, it’s necessary to keep stock in order to remain competitive on lead times. In that case, make sure it is assemble-to-order stock or, if that’s not possible, at the very least create two stocks – one for recurrent orders and the other for project orders – to avoid cannibalization of your own products. If a huge project order swallows up all your inventory, you’ll face service problems with your regular, loyal customers and that’s bad news for every type of company. By sticking to two forecasts and a separate stock policy, you protect your regular and recurring orders from the heavy impact of demand peaks, which will ultimately lead to happier customers. And isn’t that what we all want?
Curious to find out how much benefit a Collaborative Demand Planning Process can bring to your company?