FORECAST – the essential ingredient

Why Forecasting matters

Forecasting is the process of making statements about events which have not yet happened. The word forecast means different things to different people. For businesses it means future perspectives of sales, revenue, demand, and/or supply. To the public, to us, the weather forecast plays an integral part of our lives.  From planning the day to a conversation opener. Then there are investment forecasts where investors rely on economic factors such as inflation, unemployment, interest rates. We also have astrologers who predict the future via various means such as Tarot cards. Most forecasters use historical data (maybe not the astrologers!). 

It is notable that such an important aspect of our world lives in many insular silos. There is no collaboration between the different types of forecasters and little intention to learn from other approaches.  

Part Science, part Art

The random walk theory says, “the past does not predict the future”. Should we then rely on predictive models based on past data to forecast the future?  Yes and no.  Predictive models are invaluable to enable you to process vast volumes of data and variables. Their value depends on the skills of the data scientist who builds the algorithms and their understanding of the business.

Data or science-based predictive forecasts always need to be reviewed by a business realist to test their relevance. There will always be unexpected events such as Covid-19, corporate collapses or natural disasters such as a Tsunami. Best forecasts come from pairing people who have the skills of a scientist (data scientist) with the skills of an artist (business entrepreneur).

Why forecast accuracy matters   

One of my CFOs, emphasized the importance of good forecasting with a story of Apple’s demand forecast. The leadership team including the CEO and Sales VP predicted flattening demand. The CFO predicted an increase in demand. The leadership team committed to the conservative path – and when the demand increased rapidly, there was insufficient supply. Apple lost market share worth hundreds of millions.

Accurate forecasting is critical when changing models. An oversupply of products and parts just before a model change can be disastrous. Weather forecasting is critical for farmers decision-making. Earnings forecasts play a role in share prices. Interest rate forecasts play in house prices. Every forecast has a cause-and-effect play. People who provide input to forecast, build forecast or decide on the forecast, need to understand implications.

Diversity and Inclusiveness in forecasters

Diversity and Inclusiveness in forecasting means understanding the individual forecasting bias. Not everyone has the same style of forecasting. I categorize forecasters into 4 categories:

·        Scientist – A person who will focus on data quality and data models to determine the forecast. A great approach for businesses which have a large number of products, customers, or variables. There are three basic types—qualitative techniques, time series analysis and projection, and causal models.

·        Entrepreneur – A person who has the skill to identify the focus areas and then laser into the most critical such as key deals. They determine the forecast based on both quantitative and qualitative factors.

·        Sandbaggers – A person who holds good probability deals in a bottom drawer, working hard and quietly. Bringing in the deal at the last minute. Should we congratulate them? Or should they be penalised for missing the forecast?

·        Magician – A person who continues to work incredibly hard on the near impossible and pulls a rabbit out of the hat. We need them. They should be congratulated and not criticised for missing a forecast.

·        Jury of Executive Opinion. This method of sales forecasting is the oldest. One, or more of the executives, who are experienced and have good knowledge of the market factors decide on the sales forecast.

What’s next?

Many of us rely on the weather forecast from day-to-day activities to business planning. Farmers rely on forecasting to plant, harvest and prepare the land.

Despite many advances in forecasting technology, there will always be an unpredictable component. In 2004, I was in Colombo and planned to visit relatives 3 hours down a coastal route. A late decision on 26th December to leave after lunch, saved me from one of the most devastating events of modern times. If we stayed with the original plan to leave in the morning, we would have been at Seenigama when the Tsunami hit. Seenigama is where a train derailed, the largest single train disaster in the world with 1,700 fatalities. Waking up to a bright sunny morning, gave no indication of the disaster awaiting. Many people had never heard of the Tsunami. Can we predict the next Tsunami or the earthquake?

We can continue to build techniques, data models, processes for improved forecasting. And we should. What is more important than forecasting is building resilience and agility. When unexpected events happen, agile adapters are the winners. They are back on track quickly and then on the podium as a winner.