Global Forecasts of Demand in 2020

A global educational services organisation was undertaking a major strategic review which was underpinned by a vision of the world in 2020. To support this review, a well known management consultancy firm had forecast the number of products (known as programs) it could expect to sell between 2007 and 2020.

Unfortunately the forecasts produced were not felt to be credible by the senior management team. They were concerned that the forecasting methodology was unsound and decided to seek advice from an independent statistician.

Given that the organisation had been selling 4 kinds of programs in over 125 countries for over 20 years, they knew that long term forecasting would be no easy matter and thus an extensive search was undertaken.

Why the Company Came To Me & What I Did For Them

After contacting me, they quickly realised that I had both the technical expertise required as well as the commercial awareness needed to produce high quality forecasts that be understood and believed in by the management team responsible for the strategic review. On reviewing the forecasts produced by the management consulting firm, I produced a 7 point list of the deficiencies in the methods used and outlined how I could overcome these. The worst error of all was the belief that a 13 year forecast could be produced merely by extrapolating the trends seen in the last 3 years.

Another fundamental error was the failure to calculate confidence intervals in the forecasts. Without these, the organisation had no idea how much credence to give to the forecasts.

After an initial multivariate analysis of all the individual time series, I decided to extrapolate the trends for approximately 10 time series in all using a mixture of standard time-series methods and other methods I developed myself that.

These represented the key global trends in the number of programs sold. These 10 series were then dis-aggregated into the 1000 or so individual time series required.

What Happened Next?

The organisation were now in a position to disseminate my forecasts to all its managers and agents around the world to have them sense-checked. The response was very favourable and as a result, they now felt they had some figures on which their 2020 vision could be based and a strategy to achieve that vision could be developed.