Forecasting in disruptive times – what’s the point?

You wouldn’t plan a picnic without checking the weather forecast, you wouldn’t leave for the airport without checking what the traffic is likely to look like. We do this because by estimating the impact our environment is likely to have on our activities, we are able to create effective plans on that basis. When we go on holiday, or get married, we are desperate for things to GO TO PLAN. Therefore, making some predictions about what might happen is an absolute must.

August is the most popular month for weddings in the United Kingdom, with 43% of marriages in 2018 taking place between June and August. With just 2% of weddings happening in January, it was the least popular month in 2018, and over 50% of weddings take place on a Saturday. In other words, people want other people to turn up and they don’t want it to rain. It is sensible to forecast then, that a Saturday in the Summer will give couples the best chance of their big day going to plan. We can’t predict exactly what is going to happen, but we can give ourselves the best chance of success by making some logical, sensible predictions.

Why then, is forecasting such an underutilised tool in operational management?

In one particular industry, Call Centres, how much work (number of contacts), how long that work will take to complete (average handling time) and when the work is going to be required (contact distribution pattern) are all highly predictable.

Forecasting plays a key role in this industry due to the availability of an existing, simple model (Erlang) and the efficiencies that can be generated by meeting demand with the appropriate level of resource. When utilised correctly it is highly reliable, in fact, forecasting in this environment can become so accurate that it releases managers from the need to plan at all.

However, not all operations are as easily predictable. Most do not have an off-the-shelf methodology that they can use with any confidence. Forecasting as a concept is often met with comments such as “what’s the point, it’s all going to change anyway” or the ever-present “how long is a piece of string” comment.

And then along came COVID-19. A once in a lifetime economic, sociological and humanitarian disaster. At the moment, there is a lack of clarity about what the future holds – 6 months away is as much of a mystery as are two years.

So what’s the point wasting time on forecasting? It’s all going to fall to pieces…. right?

Maybe. Especially if a manager’s idea of forecasting is simply an algorithm that is going to do some calculations and spit out an answer telling them what they should do – and when forecast models don’t do that – when they give the ‘wrong’ prediction – the forecast is usually pushed to the side, never to be seen again.

This ultimately leads to forecast models being created but not utilised over and over again.

The question remains, however, if you have no forecast, how can you expect to create an effective plan? Without an effective plan, how can you expect to achieve your goals? How can you expect to be ‘less reactive’, if you haven’t made any predictions?

It is important to understand that forecasting is seldom perfect. As opposed to an algorithm spitting out exact answers, managers should approach forecasting as a model, one that can be continuously improved and modified to fit changing environments. If your forecast is unstable, inaccurate or making wild predictions, it is not because your operation is particularly unique and forecasting is impossible, it is because you have not created the model effectively.

Forecasting should be the underlying basis for all business decisions. Effective forecasting gives businesses a better chance of being prepared for the future, it gives businesses time to create effective plans to deal with forecasted future events. Give your business a chance to get ahead of the game.

At the end of the day your business is more important than a picnic – so think about forecasting and ensure it adds value.