The Lancaster Centre for Forecasting (LCF) works in partnership with leading organisations across a variety of sectors including retailing, manufacturing, services and the public sector. We provide expert knowledge to improve companies’ forecasting methods and processes, and provide insights on supply chain, customer behaviour and marketing activities. “Customer Stories” presents challenges faced by organisations we have worked with, how we tackled them and what were the achieved improvements. Projects can be set up as a Masters student project or Consulting project. For further references please contact us.
Industry sectors covered
|Banking & Insurance Sector||Pharmaceutical & Chemical Sector|
|Energy Sector||Public Service & Health Sector|
|ICT and Technology Sector||Retailing Sector|
Banking and Insurance Sector
In this highly competitive service-oriented industry, customer churn and cross selling are some of the key questions for managers. Data mining techniques allow businesses to detect relevant factors that drive customer behaviour and cross selling. Typically those organisations also have large call centres where time series methods or multivariate regression models can be used to estimate demand, enabling better planning of resources.
- Call Centre Demand Forecasting @ Barclaycard
- Describing the impact of customer satisfaction in the insurance industry @ AXA Winterthur
- Modelling Customer Churn @ AXA Winterthur
Energy consumption has multiple seasonal components and exogenous indicators such as temperature, wind direction - but also bank holidays and special days. Neural networks and advanced exponential smoothing models are able to provide accurate forecasts for such problems.
- Econometric models building for British gas call centre @ British Gas
- Novel methods and empirical comparison for Danish short term gas consumption forecasting @ Dong Energy
ICT and Technology Sector
Fastpaced technology development cycles challenge the industry. Since infrastructure is expensive, being able to forecast the adoption of new technologies/services allows businesses to invest the right amount into infrastructure and innovation to fulfil the expected demand. Relevant forecasting techniques include, for example, diffusion or choice models.
- Forecasting engineer hours @ BSkyB
- Forecasting ethernet ceases @ BT Openreach
- Forecasting new acquisition calls @ BSkyB
Manufacturing organisations have to deal with demand volatility. The increased volatility, often associated with the bullwhip effect can cause substantial costs in terms of inventory and bound capital, but also reduced service levels. Accurate forecasting is crucial to mitigate this uncertainty. Other challenges include predicting and managing customer complaints, for example in terms of adequate staffing levels. Another challenge for manufacturers is forecasting the demand for spare parts. The diversity of forecasting challenges in this sector requires a wide range of forecasting methods and tools that will lead to desirable improvements.
- Forecast model selection in FMCG supply chains @ Beiersdorf
- Forecasting for spare parts by TecCMI @ TMD Friiction
- Modelling consumer complaints in FMCG @ Beiersdorf
- Purchasing new forecasting software @ Beiersdorf
- Trend estimation in the tactical horizon @ Beiersdorf
Pharmacutical and Chemical Sector
The complex production pipeline requires businesses to forecast long-term demand in advance to optimise stock levels and investment. Change-detection can lead to significant cost savings identifying changes in buying behaviour, as well as the impact of marketing activities. Models include traditional forecasting models such as exponential smoothing or ARIMA but also statistical modelling to detect changes in sales behaviour.
- Change detection in pharmaceutical demand forecasting @ Boeringer Ingelheim
- Forecast model development in supply chain management @ Bayer
- Modelling consumer complaints in FMCG @ Dow Chemical Company
- New product forecasting @ Dow Chemical Company
- Spiriva prescriptions estimation @ Boeringer Ingelheim
Public Health and Service Sector
Running hospitals and medical practices involves high costs and resource allocation is important, requiring accurate forecasts. The Centre has also worked on providing insights on the salary drift of medical personnel. These are often highly political decisions and require scenario based models. Other public sector areas involve traffic forecasting.
- Forecasting patient activity @ NHS
- Forecasting pay drift @ Department of Health
- Short term forecasting @ Inrix
This sector typically deals with high volatility in demand due to events such as promotions or weather. Another challenge is to deal with the high numbers of items that need to be forecasted. Typical instruments involve automatic statistical forecasting, promotional modelling and hierarchical forecasting. Moreover, data mining techniques enable businesses to study buyer behaviour to improve shelving and promotional timing.
- Choosing a right forecasting software @ McBride
- Forecasting for Far East @ Wilkinson
- Forecasting product returns @ Shop Direct Group
- Fruit & veg forecasting and replenishment @ Booths
- Promotional causal forecasting: A case study @ Morrisons Plc
- Setting up forecasting process for stock control @ SMD Home Ltd.
- Standard deviation and safety stock analysis @ Morrisons Plc
- The impact of weather variables on demand forecasting @ Tesco