One of the key insights from my work on business forecasting (as set out in my book Present Sense) was that if you can’t define what success looks like then you shouldn’t be surprised if you fail. 

 

But having defined ‘success’ for business forecasting it was the pursuit of an answer to the follow up question: ‘how do know if we have achieved it?’ that led me to developing a completely novel approach to measuring the quality of forecasts.

 

It soon became clear that this approach would add most value to short term forecasting of demand for the supply chain because of the number of forecasts that need to be produced and the clear and direct link between the quality of forecasts and impact on the business – the level of stock and the quality of customer service. The size of the potential benefits is simply enormous.

 

I co-founded CatchBull to develop software to exploit this technology and over time the concepts and techniques were refined and improved with the results being published in Foresight, the peer reviewed practitioner journal of the International Institute of Forecasters. As a result of their impact, I was voted into the Foresight Hall of Fame in 2022.

 

 At the heart of this thinking is that the ultimate measure of forecast quality is the amount of value that the process adds to the business, rather than abstract measures based on simple error measurement. 

 

The best way to measure this is to compare actual forecast error with the naïve error – effectively using the prior period’s actual in its place thereby playing the same role as a placebo in the clinical trial. And like in drugs development the ability of a forecasting method to outperform the placebo is the ultimate test of its efficacy.

 

One of the things that became clear to me was that the academic community that produced new forecasting methods had little understanding of the business problems that needed to be solved. And on the other side of the fence that forecasters in business and the people that led them often had rudimentary technical skills and knowledge. As a result, that they were easy prey for those software vendors who claim to have a ‘silver bullet’. 

 

In order to help close this gap I wrote The Little Book of Operational Forecasting. It is published in the same small format as the Little Book of Beyond Budgeting, but in order to make what could be intimidating content more accessible, it is written as a series of short ‘lessons’ on what to do or not to do each of which is accompanied by a hand drawn illustration.

 

The content is grouped under 5 headings:

​

• The purpose of forecasting

• Understanding demand

• Forecasting methods

• Understanding forecast performance

• Managing forecast performance

​

The ‘Little Book’ regularly appears in independent lists of recommended reading in this area. You can buy it through normal retail channels, direct from the publisher or from my bookstore in physical or PDF form.