Skip Ribbon Commands
Skip to main content

Improving your Business Forecast

14/06/2013


​The benefits of anticipating the future are obvious. From changing the way you manage inventory to customer service and order policies, almost every business practices some form of forecasting to increase business efficiencies and meet customer needs. One of the basic rules of business however, states that forecasts tend to be wrong and are at best, somewhere between ‘lousy and lucky’ in their predictions.

Rather than conceding defeat to the challenge of accurately predicting the future, it is possible to improve the accuracy of your forecast and better manage uncertainty. The reward for this effort is significant cost reduction and improved customer service.

Assign a Chief Forecaster

Like many things in business, if you want it done right, make someone responsible for doing it. Assign a Chief Forecaster, and have them gather forecast data, analyze the data to create realistic predictions and continuously update and communicate the forecast key stakeholders.

Ideally, this person is an insightful veteran of your organization and industry and is adept at distilling the truth and identifying trends from sales forecasts, marketing promotions, new product introduction schedules and a range of other dubious inputs. All of these sources need to be considered to provide an informed guess of what the future really holds.

Break down by service or product line

Your forecast will be more accurate if you forecast by product or service family. In this instance, a family is a logical group of products or services that share some common characteristic (materials, customer, uses, etc.) Often, knowing the forecast for each of, say, 25 products within a family is information overload. And while the overall forecast for the family is likely to be reasonably accurate, the forecast for each of the individual 25 items will be far less accurate.

The rule of 3

As we’ve already established, all forecasts are ultimately wrong, so it makes sense to include three numbers when forecasting. These three numbers should include your best guess, along with the worst and best cases. Preparing to provide your customer with something between 95 and 105 widgets is very different than preparing to provide something in the range of 50 to 150 widgets, although both scenarios might appear as a forecast for 100 widgets.

Reduce internal lead time

As you work toward improving the accuracy of your forecast, you’ll also want to work at becoming less dependent on your forecast. For example, if you could buy all required parts, build product and ship orders within five days and your customers only expect delivery ten days after ordering, you would not need a particularly detailed or accurate forecast (or maybe any forecast at all!) Even a modest reduction in the length of time required to deliver your products and services will make a big difference in the ‘accuracy’ of your forecast, since the early part of a forecast is almost always more accurate than far-ranging forecasts.

Of course, reducing internal lead times is also a great way of driving down cost (to speak to another law of business: the longer it takes to build a product or to deliver a service, the higher the cost).

So go ahead and add ‘improve forecast accuracy’ to your list of key business improvements for this year. You might need to deflect a chorus of “that’s as good as it gets,, or “our industry is different,” and other common excuses for lousy forecasts, but these efforts will go a long way toward improving the performance and output of your organization.