The importance of being state-of-the-art in forecasting
02 October 2014
02 October 2014
Investing in your forecasting adds value to more ways than just more accurate forecasts. Being critical for decision making, forecasting can become a competitive strategic advantage.
The Lancaster Centre for Forecasting over the years has engaged with several companies from various sectors to support them on their forecasting challenges. As diverse are their forecasting problems, equally diverse are the different approaches and attitudes companies have exhibited towards forecasting as a business function.
Nowadays, companies have increased access to historical data and external drivers, yet these are often not used in practice. There are several reasons for this, but I will focus on two here: capabilities and expertise. Often organisations perceive forecasting as a necessary function, with its sole purpose being to produce some figures that are fed into the planning process. The quality of the forecast is not evaluated and crucially there is little consideration whether a forecast is as good as it could potentially be, given the available information. This is typically reflected in the software that supports forecasting, which most of the time is not specialised, rudimentary and sometimes even inappropriate. Therefore, there is an evident lack in terms of capabilities.
Often limited capabilities are related with lack of expertise in forecasting. There is ample evidence in the literature that human experts add value to forecasts and successfully incorporate useful information in their predictions, albeit in an inconsistent manner. However, there is also overwhelming evidence that accurate and reliable statistical forecasts are fundamental to achieving good results. Therefore, in principle forecasters need to be experts both in their markets but also in forecasting as a discipline. Obviously, such forecasters can use appropriately all available information, produce accurate forecasts and drive an increase in capabilities, both in terms of processes and software.
Why is this important? In a nutshell forecasting, especially nowadays with the increased availability of information and computational power, can be a competitive strategic advantage. Forecasting is not to merely provide figures that are fed to the next company process. Predictions are the cornerstone of any decision making. Therefore, forecasting is crucial for identifying and capitalising on market opportunities, providing business intelligence and helping you understand your market and consumers.
Let me illustrate this with a real example. Sky, a private UK media company, has been working together with the Lancaster Centre for Forecasting for more than a year to improve its forecasting processes and expertise, with the aim of being state-of-the-art in forecasting and therefore capitalising on the opportunities that this brings. We have worked closely with the S&OP team that is responsible for planning of engineer-hours and call centres that handle sales and support calls. Sky S&OP team invested heavily in their forecasting training, in both areas of time series modelling and causal regression forecasting. These gave them the ability to standardise and automate a substantial part of their forecasting and start evaluating how to further improve it with the inclusion of additional marketing and market variables. The latter in particular has given forecasters the ability to start quantifying the effect of various BskyB decisions, such as promotions, and competitor actions, thus being able to feed this information back to the organisation to optimise operations, cut costs and capitalise on opportunities. A reflection of the expertise that the team has achieved is that they are now at the final stages of being certified by the International Institute of Forecasters, the leading international forecasting body for academics and practitioners.
In my opinion, a further major advantage investing in the expertise of the team was that BskyB recognised current limitations in software, processes and current forecasts and are quickly proceeding to rectify these. This demonstrates that the team has reached a critical level of expertise to internalise forecasting knowledge and start driving innovation. It is fair to say that we have been impressed by their effort, energy and skills of the various forecasters in the S&OP team, throughout our interactions during the course and further discussions.
The forecasting team retains close ties with the Lancaster Centre for Forecasting, evident in two very interesting MSc student projects that just concluded. Such activities allow them to further increase their expertise, working together with students who have been formally trained in the latest forecasting and business analytics trends.
The team has clearly benefited from the process, improving and steadily moving towards becoming a competitive strategic advantage for Sky. It is my hope that I will see such teams increase in size and skills, bringing state-of-the-art forecasting to businesses.
The Lancaster Centre for Forecasting is always interested in collaborating with industry, bringing our forecasting research and expertise to businesses. For more information please contact LCF or Nikos Kourentzes.