In today’s environment it’s not good enough just to use last year or last year + x% to base the following year’s business plans on, the landscape is changing to rapidly to rely on past performance with an expected growth, and when it comes to operational planning, just achieving or exceeding historic sales trends at an aggregated level leads to increased cost and inefficiencies throughout the Supply Chain at the item level with the potential of customer service failures. A more scientific view needs to be considered when generating and managing the Demand Forecasting and Inventory Planning process as a whole.
While creating and maintaining a meaningful Demand Forecast can seem a daunting and time-consuming task when we know that the past isn’t always a guide to the future, breaking the numbers down to the key elements makes it much more manageable.
Demand Forecasting is made up of 3 key elements:
· Base – demand that is not influenced by planned activity i.e. promotions or targeted marketing.
· Activity – planned activity intended to increase sales or market penetration.
· Events – events that occur outside of business control that influence sales levels, i.e. Easter (moves each year), sporting or industry media events etc.
Compiling the Demand Forecast for budget purposes is a different process to maintaining the forecast for operational purposes. We know that the further out we try to predict the less accurate we are likely to be, so compiling an annual volume at a granular level doesn’t always add value.
Compiling the budget numbers from an acceptable bottom-up forecast over the business forecasting horizon, with a scientific view of past trends, market intelligence, future activities and growth expectations, and that are measurable over time against attained volumes is typically sufficient.
Equally, it is not just about exceeding the budget numbers. While the objective is to grow the business and exceeding the budget numbers is a good thing, it needs to be done in an effective and manageable way to maintain other targets like Customer Service Levels, Operational Efficiency and Financial stability.
Once the budget is set and agree there is something to measure against however, maintaining the operational forecast is not just a once-a-year process, it needs to be continually maintained both with short-term (at least to longest purchase/supply lead times) and with mid/long-term changes, typically managed through a Sales & Operations Planning process (S&OP).
Maintaining changes to the short-term Demand Forecast provides a view of risks and opportunities, i.e. is the volume achievable within the capacity and supply constraints or do we have an opportunity to achieve additional sales? Maintaining the forecast over the mid/long-term horizon ensures the business always has a view of Year to Date variance, Year to Go expectations and therefore full year financial projections against the original budget. This allows stakeholders to make the necessary decisions to mitigate issues or exploit opportunities.
However the forecast is compiled, measuring the accuracy and bias is needed to provide a benchmark and while there are various industry standard benchmarks reported it’s not always easy to find one that relates to a specific sector. Measuring and understanding Forecast Accuracy and BIAS levels specific to an individual business is more important as it provides a benchmark to improve against, and if you manage forecasts from various areas of the business, i.e. Sales, Marketing, Finance, Customers etc. it provides a comparison and confidence level to work with.
Forecast Accuracy is typically measured using MAPE. While MAPE (Mean Absolute Percentage Error) is technically an error measure the inverse equates to accuracy. Typically, MAPE is measured using a lag, i.e. actual orders vs. the forecast set 1 month in advance (the lag allows for a reasonable supply lead time rebalancing) and is aggregated up from a SKU level measure to category level. Aggregating up takes out some of the noise while still maintaining the detail level accuracy and providing a good representation of category accuracy and bias.
There are various methods of compiling and maintaining the business Demand Forecast.
ERP systems usually provide functionality to generate a forward Demand Forecast, but the functionality is normally limited to some basic options like using last year’s sales volumes or applying some simple growth metrics, and while these will give a basic demand profile to work with it doesn’t consider any growth or decline curves, changing annual events or any sales or marketing activity planned to grow sales. Using this demand profile risks generating an inaccurate financial budget and operational supply plan.
Alternatively, excel demand forecasting tools are commonly maintained. Spreadsheets have seemingly infinite flexibility to allow users to create a bespoke demand forecasting system tailored to specific business needs, they have good graphical capabilities to easily review the demand forecasts and are easy to update, maintain and share. The downside is that reasonable spreadsheet skills are required to get the best out of them. Read our blog ‘Are Excel Planning Tools an Effective Option?’.
As ERP systems don’t provide the functionality to generate and maintain a meaningful forecast, and spreadsheets, while allowing more flexibility, require technical skills to develop and they can be time intensive to maintain, more sophisticated and user-friendly demand forecasting solutions have emerged.
As more emphasis over the years has been given to compiling and maintaining the demand forecast to ensure the starting point for all business financial and operational planning is as accurate as possible, more cost-effective specialist demand forecasting systems have become available.
Specialist Demand Forecasting system use sophisticated algorithms to analyse sales, usually over a minimum two years of history, to generate forward demand profiles and provide the functionality to maintain and compare collaborative forecasts, compile event and activity volumes and support business reporting needs, and in turn take out the stress of the annual task of compiling the budget.
With the advancements in the way software has become available over the years i.e. cloud based Software as a Service (SaaS) where tailored solutions are provided on monthly licence agreements without incurring high capital expenditure, the solutions have also become much more affordable and provide much greater functionality and reporting capability that simple spreadsheets.
From the outset of planning for Brexit then managing through the Covid pandemic there has been a lot reported by industry institutions and media outlets of the challenges businesses have faced and the changes required to build more resilience in to supply chains – read our blog ‘Supply Chain Challenges, what changes the experts are suggesting’ which reviews the suggestions made by CIPS and the BCI.
In our recent survey the main goal of Supply Chain Management is to add value to the customer, equally 60% of participants report forecast accuracy of less than 75%. Can value be added to the customer and the business operational and financial targets be achieved with forecast accuracy less than 75%?
To take part in our survey and click https://hms.avercast.eu/
Comentarios