Fancy tools won’t make up for a bad forecasting strategy

A lot of business planning depends on forecasts and in the quest for ever more accurate and reliable forecasts, a lot of people turn to increasingly fancy tools and algorithms.

But what if the problem with your forecasts isn’t the tools you are applying, but your whole approach to forecasting in the first place? If your basic strategy is wrong, no amount of fancy processing will improve the quality of your forecasts. In fact, it could make things worse, by masking the incorrectness of the data and giving a dangerous false sense of confidence.

Of course no prediction of the future can ever be better than an educated guess – but we can improve our guesses by improving the quality of the information we base them on. It’s as the old adage goes: Garbage in, garbage out. If your forecasts for next year are based on nothing more than this year’s figures, their value is low.

The most critical question to ask before you embark on any forecasting exercise is this: How much can we expect next year to be just like this year? Close on its heels should come the next question: How many things that are going to make next year different do we already know about, or can we find out? There will always be unknowns lurking around the next corner, but it makes sense to reduce their number as much as possible.

The question about what might change for your business next year is not one that can be answered by the finance department alone. Sure, there are some high-level variables you can get from professional economic analysts; but when it comes to the very specific issues that might affect your business, you need to be turning to your own managers and operational staff.

One can’t help wondering, for example, if whoever compiled Lonmin’s forecasts for this year was aware of the fact that a rival union to NUM was organising at their Marikana mine. If they had known, might they have factored in an increased risk of industrial unrest?

An easy first step is simply to ask the question: “Is it reasonable for us to assume that you’ll match this year’s production next year?” If the answer is no, seize the opportunity to find out more. What does your sales staff know about what competitors might be planning? What do your production managers know about a supplier that’s facing difficulties?

Instead of using technology to put your numbers through fancy contortions, use it instead to solicit information from your staff, quickly and efficiently.

It’s a bit more work to do this research than simply to run this year’s figures through a clever algorithm, of course – but the benefits go beyond more well-founded forecasts. If people know that their input has been heard, they’re typically more inclined to commit themselves to turning those forecasts into reality – and to contribute intelligence to next year’s forecast in turn.