A Five-Step Approach to Effective Demand Planning Implementation
Good handling of market demand data is one of the most vital concepts in any supply chain. The correct management of demand information can greatly influence the level of integration and responsiveness, and has a direct impact on customer service and inventory levels. If customer demand is the activating element in the supply chain, it’s quite clear that a multi-step operational supply chain management process to create reliable demand forecasts can play an active role in improving supply chain effectiveness. But how to create powerful demand planning implementation? The five-step approach outlined below provides guidance.
1. Start with a demand analysis
It’s an old saying: ‘garbage in, garbage out’. Any forecasting process starts with analyzing and understanding your sales history and cleaning up your dataset. In addition to the typical data cleansing tasks, it’s also important to establish focus for the forecasting process: which items do we need to review?
Doing a customer-product segmentation prior to the demand planning implementation is a great help in this respect. Additionally, opting for a separate forecast for the recurring items and promotional, project or tender items will sharpen your focus. High-volume, fast-moving items that can be forecasted accurately based on statistics should be left to the automated forecast of the demand planning tool. However, you should always incorporate an exception-based review list.
2. Rely on a quantitative baseline forecast
Statistics should always be your starting point rather than the finishing post. This not only makes life easier for Sales, but also helps to remove bias from the forecast because people have a natural tendency to over-forecast. Furthermore, it will free up extra time for your sales team to focus on lumpy and erratic figures.
At the same time, you should be very careful with statistical forecasting. In general, the forecast statistics are too complex and often poorly understood. If you have an effective process and a good tool, statistical forecasting can easily generate 30-40% more efficiency. If the results applied incorrectly, however, the only outcome will be frustration. So look for a knowledgeable partner who can help you select the right level of complexity for your business and ensure you benefit from improved performance as a result.
3. Strive for a collaborative demand planning implementation process
We can’t over-emphasize the importance of gaining extra input for the statistical forecast from the right people, both from within the company and from key customers and distributors. Demand can be influenced by a myriad of factors, and way more than any statistical model can currently handle, so, statistics alone can’t do the job!
You have to rely on a collaborative process within and beyond company boundaries to receive the additional, crucial demand information such as about product launches, substitutions and end-of-life products, for instance. Other types of relevant information include promotions, price changes and marketing campaigns, and projects and tenders could also be a disruptive factor in your supply chain. This extra demand information is typically managed via reviewing lists. Besides these, it can be very useful to ask your sales team to validate the statistical forecast, which is often easier to do on an aggregated level.
But it doesn’t end with your sales team’s contribution. You should stop trying to guess what your most valuable and important customers will buy. Instead, go and talk to them and set up a collaborative forecast approach. Do the same for your other important channel partners too.
4. Invest in performance management
The numbers tell the tale! In return for all your efforts to improve the forecast, you should also be able to measure your progress. Reducing the forecast error will improve service while lowering cost and inventory, so it’s worth ensuring that the forecast accuracy is monitored effectively.
In a typical situation, Supply Chain will generate the statistical forecast, Marketing or Product Management will add information about promotional campaigns, new product launches and end-of-life products, and Sales will add customer data. Ensure that your tool enables you to check the added value of each of the different forecast versions and include a feedback cycle.
At the very least, follow up on these three key metrics: the mean percentage error, the mean absolute percentage error (MAPE) and the stability of the forecast accuracy.
5. Hold a demand review meeting
The final step in demand planning implementation should be a demand review meeting, led by the demand manager or the S&OP manager. Supply Chain is the only department that can be regarded as being ‘neutral’ with regards to Sales, Operations and Finance and hence is in the best position to chair demand review meetings.
Some tips and tricks for a successful demand review meeting include: avoid getting into discussions about the data, don’t allow the S&OP forecast to differ from the financial one, shift the discussion onto values instead of volumes, keep the meeting brief and powerful, and prepare scenarios in the case of uncertainty. This will all help you to build consensus and ensure a high-quality forecast in the S&OP process!