Demand Forecasting & Planning: Improvement Programs
to the Forecast Solutions website
Training Courses

Forecasting in Excel

Forecasting Service

Pricing Models

Weather Effects


Improve Forecasting Process and Methods to boost Forecast Accuracy

Sales forecasting is a critical process that affects business and supply chain planning throughout the entire company. Forecast Solutions is well qualified to undertake an end-to-end review of all forecasting methods and processes with the ultimate aim to improve forecast accuracy.  On a more modest scale we are happy to help with any individual component of the forecasting process.  Software is often important, particularly in large or complex systems.  We can help ensure best use is made of any existing forecasting software, but can also recommend new tools if such are deemed to be needed.  If company resource is limited we can provide a temporary or permanent service to help with regular updates in your forecasting system or can supply a full outsource of the forecasting process.

Understanding the current state

The first stage of any improvement program is necessarily to understand the current sales forecasting process in terms of its scope, forecasting methods and operation.  As well as the mechanics it is also good to be aware of the considerations and events that resulted in the current state and the views of key stakeholders on prevailing strengths and weaknesses. 

As the main objective of the forecasting process is to maximise forecast accuracy, currently achieved accuracy should if possible be calculated on a meaningful basis.  It may be beneficial to set up a separate forecasting database to facilitate a comparison of existing forecast errors with those that can be achieved by deploying a good standard set of statistical methods.  At the end of the day the final set of forecasts made after the application of commercial judgement should be superior to statistics alone, but a statistical check can be very helpful in identifying weaknesses in current methods and process.

Identify and test improvements

When a full understanding of the current state has been reached, supported by appropriate documentation, all parties should then step back and critically assess the current system. It may be that a process installed with due diligence some years ago no longer meets the full business need.  For example, a system designed to support short term planning might not adequately fulfil the needs of medium term planning activities such as sales and operations planning.

Obviously there are practical considerations including the often large investments that may have been made in forecasting software, but software should ideally be regarded as a tool to support the sales forecasting process, not something that defines the process.  Ideally we should think in terms of a 'forecasting support system', not just a collection of statistical tools.  At the end of the day it may be necessary to consider additional or replacement software.

Potential improvements should be discussed and prioritised.  Of great importance is the organisational and procedural framework within which sales forecasting will be carried out. Effective forecasting usually involves many individuals both inside and outside the business and the necessary collaborations must be supported by the forecasting process.  Proposed changes should ideally be tested before implementation.

Implement and monitor

The agreed changes should be implemented in line with a realistic timetable. Appropriate forecast accuracy measures should be put in place and regularly monitored as part of a drive for continual improvement. When setting KPIs it is best not to worry too much about industry benchmarks, rather to focus on accuracy currently achieved, put in place challenging but realistic targets and strive to improve forecast accuracy continuously over time.