Earlier this year I was involved in a talk with a conglomerate on inventory management and processes on how we can bring inventory levels down. Not surprisingly the forecasts or rather the inaccuracy of forecasts was a big bone of contention for all the participants and how it resulted in high non moveable inventories.
The reality is there is no known magic for making forecasts accurate. There are obviously methodologies to help close some of the gaps both during the S&OP process and also during executing plan. As a subject of a fruitful discussion, here are a few tips we can use to manage forecasts which can actually results in some BIG WINS for inventory reductions.
Many organisations do not practice this. For those who have been getting away with sub- standard forecasts, it could mean a clear message to start improving. Strive to have good documented process in place. Measuring forecast accuracy is one of the most important steps to improve the forecasts.
The trick here is for Demand Planning to get close as possible to closing the gaps. At the very minimum establishing targets for forecast accuracy for the items at product level, product groups and even product lines will help a lot.
It is very difficult to foresee what can happen 6 to 8 months or even 1 year down the road. One variability we can proactively try to overcome is raw materials lead-times. Since forecasts at the very least need to cover lead times of raw material delivery, pushing to reduce raw material lead times can further help support accuracy.
Very important to have processes in place that manages all Bill of Materials and Build of Routes with accuracy. Inaccurate or outdated BOMS can result in wrong materials being ordered.
While this may come with added advantages like price reductions, be also on lookout for any non-cancellable clauses that can be attached to this. Potentially we can end up with high inventory holdings we do not need.
Not sales forecasts. The key to improve forecasting accuracy is looking at sales forecast and demand forecast as 2 different entities. As an example when a company runs out of stock on a product, it means no sales for that particular product despite having demand. However based on the low sales history, the forecast generated using sales history could end up showing a very low demand and so misinterpreting the real need.
Instead of going through the entire large portfolio of items in detail, demand planners should filter the data that is relevant and of importance and act on that first. The best way to do this is categorize inventory based on an ABC classification of what is most important.
The need for a process between product terminations with rest of the supply chain. A disconnected process has often resulted in excess finished goods and also raw materials.
There are more ways to skin the cat and these are some just of them. We know why forecasting is important and actions that needs to be done to improve accuracy. Do leave your comments below as I would like to hear from you.