How to Create Accurate Sales Forecasts

By Laine Chroust Ehmann

Sales managers typically view forecasting as an exercise in frustration and fortune telling. Why? Mitchell Gooze, author of The Secret to Selling More, says creating accurate forecasts in the traditional way is tough work for three reasons:

1) Customer behavior is difficult to predict.

2) Managers are dealing with sales teams full of personalities that are largely optimistic and inclined to give the best possible spin on the story (and these team members are continuously asked to increase the amount they sell).

3) Salespeople must predict not only whether or not the customer will buy but also when the order will come through.

As a result, most forecasts deviate significantly from reality. Gooze’s solution: salespeople must recognize that buyers make purchases according to their process and timeline, not the selling company’s.

“There’s a belief that salespeople can ‘make something happen,’ whether or not the customer is ready to buy,” says Gooze. “Salespeople tend to match the customer to the selling process, but the real question is, where is the customer in the buying process?”

Recognizing that you’re not in the driver’s seat will actually allow you to increase the accuracy of your predictions. “Just because customers are in control of the buying process doesn’t mean their behavior isn’t somewhat predictable and logical,” Gooze explains. Once you identify the customer’s buying process, you can match your selling process to it and then create a forecasting model.

Another key to greater accuracy is to look at the probabilities of moving from step to step in the process, not of closing the sale. Because each step in sales (and buying) depends on the previous one, the outcome of this stepped approach is more accurate than simply saying, “I think 15 percent of our active prospects will close before the month’s end.”

Additionally, over time the entire model becomes more precise as you replace estimated probabilities with real ones. For instance, you may estimate that for each RFP (request for proposal) you respond to, there’s a 20 percent chance of being asked to give a formal presentation in the next two months. Eventually you may discover that the chance is actually closer to 10 percent. You can then replace estimates with real data in your model, thereby increasing your forecasting accuracy.

Accurate forecasts often mean lower numbers as you remove the puffery and unfounded optimism, but reality is vastly preferable to a dream world. “The purpose here is accuracy, as opposed to delusion,” says Gooze. “This forecasting method forces you to identify what’s going on.”