Effective demand forecasting: beyond improving forecast accuracy

“The trouble with weather forecasting is that it’s right too often for us to ignore it and wrong too often for us to rely on it” (Patrick Young). In general, when you improve forecast accuracy, it brings value in terms of improving KPIs and reducing manual intervention, however it is not (yet) ready to fully replace the manual forecasting process in large multinationals.

Aligning the way of working

Planning Services, the business process outsourcing service of EyeOn, offers robust, recurring outsourced services focused on high quality forecasting, inventory optimization, and actionable end-to-end supply chain insights. One might think that the main goal of an advanced forecasting service is to optimize the forecast generation process to improve key performance indicators such as forecast accuracy and bias. However, in the first few months (or weeks) of collaborating closely with a customer this is usually not the highest priority. Together, we decide what the monthly (or weekly) service should look like and who should be involved. For multinationals often the way of working across the business units can be quite different, hence the first step is aligning the way of working across all units and locations.

Improving forecast accuracy

Second, high emphasis is put on improving the customer’s (master) data quality and creating a persistent storage of the important input and output datasets each cycle. Persistent storage is any data storage that retains after power to a device is shut off, whereby we guarantee no data is lost. This is key to running a robust recurrent service. Third, we navigate the main stakeholders through the agreed process step by step. Optionally, we provide in-depth trainings on the statistical forecast generation concepts and/or the best forecast enrichment approach.

Focus on change management

April 21st, 2022, we hosted an online seminar about advanced forecasting enhanced with a planning enrichment tool in Jedox. During the seminar, one of the participants asked us: “From your past experience, have you encountered resilience to change? How do you deal with that?” In each of the steps described above, we take the customer along on the journey. The general approach is to first make sure the advanced forecasting concepts and theory is clear to the stakeholders. Thereafter, we ask the stakeholders to search for examples of forecasts they did not expect and/or for which they have certain business knowledge. Together we define next steps, i.e. how to adjust the forecasting logic to allow for these exceptional cases, if relevant.

This expands the statistical forecasting knowledge of the people dealing with the forecast, which increases the buy-in and has proven to steadily improve key performance indicators over time. Simultaneously, EyeOn Planning Services takes a clear continuous improvement approach. For example, we review the settings in each step in the statistical forecast generation engine at pre-agreed intervals.

This mutual stepwise approach brings the process to a higher level, while enabling our customer to eventually bring it back in-house.

Read more about our approach to forecasting and demand management