How to Improve Sales Forecasting

Sales forecasting has changed with the popularity of sales force automation and with CRM moving to a software-as-a-service delivery model. Here, Knowledge Center contributor Joe Ruck explores the art of sales forecasting, including how sales forecasts are often erroneously based on the opinions of individual sales reps, who rarely know without a doubt whether a particular deal will close in the designated time frame.


Fans of the Fox television series "House," a medical drama starring cynical but effective Dr. Gregory House, will recognize his catchphrase, "Everybody lies." Like other good one-liners, it rings true beyond its original context. One area where it has particular resonance for me is in sales forecasting.

Sales forecasting is serious business. Financial markets expect public companies to deliver a smooth revenue ramp. A missed sales forecast carries dire consequences. Wall Street will punish "your miss" by hammering your company's stock. Of course, enterprises need to get their sales forecasts right for their own reasons. If done well, a sales forecast helps assure that there is sufficient capacity to process orders and ship products on time. Get that wrong and you're saddled with either expensive inventory or lost sales.

Extrapolation versus forecasting

Forecasting remains more of an art than a science. Even in large companies, sales forecasts are often based on the opinions of individual sales reps. Therein lies a problem. Only rarely do reps know without a doubt whether a particular deal will close in the designated time frame. There is a great deal of unpredictability that accompanies the purchase of consumer goods, and it's worse in sales to businesses. Most business-to-business transactions involve multiple influential decision makers and a decision process in which group dynamics play an important part. Under those circumstances, a sales rep's capacity to predict an outcome is often not much better than relying on the law of averages.

A good substitute for individual deal forecasting is extrapolation. Something like, "Last quarter, I had $1 million worth of opportunities and $500,000 came in. This quarter, I have $1.2 million worth of opportunities so, hmm, let's say $600,000 will be the forecast."

Hopefully, if you have 100 sales reps, the statistics all come out in the wash. But the sad fact is, statistics are only as good as the data on which they are based. Follow them blindly and the result will be no better than a simple guess.

Sales managers: Get involved

Companies are under tremendous pressure to meet their sales targets. This pressure gets transmitted directly to the sales force in the shape of both the carrot of huge commissions and the stick of dismissal. The thinking is that this approach will lead to better forecasts. However, this is not the case. In particular, the threat of termination, coupled with the inherent uncertainty of forecasting, turns what should be a tightly controlled and transparent business process into an opaque game of hide-and-seek.

The problem is that if sales reps were to forecast anything less than say, 70 percent of their quota, they would put themselves at risk of being terminated. On the other hand, if reps forecast numbers over 130 percent, they risk having their territory reduced. So, it should be no huge surprise that all reps feel the need to forecast in a tight range, along the lines of 80 to 110 percent of quota. Everybody lies.

Over-allocating quota is common practice

Experienced sales vice presidents understand this and attempt to compensate by over-allocating quota. This way, at each level in the sales management chain, you can still commit to 100 percent-even with your reps committing an average of 95 percent all the way up to the executive suite. Most of the time, this method makes the numbers balance out.

In this scenario, it is the sales manager's job to peer through all the fog and try to build a picture of which deals are really likely to happen and which are not. This will be done via a process of cross-examination: "Do you know the name of the decision maker?" "When was the last time you visited the customer?" "Show me the most recent e-mail reply from the prospect." "Can you set up a meeting for me to meet with the customer?" The cycle repeats itself, since the same logic forces the sales manager into delivering a forecast bounded to 80 to 110 percent of quota to his or her immediate superior, and so forth, all the way to the top.