Fast-moving consumer goods

Forecasting is a key driver to improving to supply chain performance, in particular for fast-moving consumer goods (FMCG) companies who need to forecast hundreds or thousands of products automatically.

Assortments are often heterogeneous, with different time series patterns of seasonality, trends and level shifts from listings & de-listings across product groups and regional markets, which require robust tools for automatic forecasting model selection. External influences from out-of-stocks and promotions need to be captured automatically in advanced statistical algorithms or through aided expert judgment. The recent shortening of product-lifecycles, proliferation of promotional activities and the increasing availability of downstream data are demanding more sophisticated forecasting solutions than ever before.

Despite being at the core of many standard ERP and advanced planning systems (i.e. SAP APO-DP, JDA, Oracle, Demantra, Manugistics or I2) the forecasting functionality is often little understood and poorly customised, and allow for significant gains in forecasting accuracy, reduced safety stocks and higher service levels when properly aligned with the product assortment.

The Lancaster Centre for Forecasting has developed specialised solutions, providing support and tools for improving statistical baseline forecasting:

Auditing image

Auditing forecasting and assortments

LCF provides a structured analysis of assortments by importance (A-B-C), forecastability (X-Y-Z), product lifecycle (New-EOL) and time series patters (Seasonal-Trended-Intermittent/Lumpy-Discrete).

Statistics image

Improving statistical baseline forecasts ‌

Advanced automatic model selection (e.g. for exponential smoothing) outperforming standard approaches in systems by up to 30% forecast error (MAPE). These are available as one time forecasting audits of systems, methods and processes, which provide offline reports for the best forecasting methods for each SKU-location combination, to advanced customisation and integrated tools for bespoke automatic model selection custom made for a company’s unique product mix, or specific forecasting horizons.

  Success story Modelling consumer complaints in FMCG  @ Beiersdorf
Success story Trend Estimation in the Tactical Horizon‌ @ Beiersdorf

Enhancing expert judgment for S&OP‌

Enhancing expert judgment as well as developing robust processes to combine expert judgmental overrides with baseline forecasts in S&OP processes.

Integrating downstream customer demand, sales and promotions data‌

Integrating downstream supply chain information from distribution centers, 3PL, wholesalers, retailers, and ePOS data from consumer sales in planning.

In-house/custom-made training courses

Since 2002: Beiersdorf AG (20+ courses in Germany, Netherlands, Brazil, Singapore). 

Trained over 200 demand planners in statistical forecasting with SAP APO DP, using hands-on exercises in Excel & SAP APO-DP using real time series.
2002: Bayer Cropscience
2001: Sanofi-Avents
1999: Celanese
Bayer Business Services

For any enquiries please contact consulting@forecasting-centre.com.