How to Close the Gap Between Sales Forecasting and Reality

By Dave Boyce On May 13, 2019

Sales forecasting is part of every company’s standard operating procedure. Each sales leader has their own process for arriving at a forecast, but the end result is the same– sales leaders live and die by the forecast.

There is nothing more frustrating than sorting through the mountains of data that get rolled into a forecast. Each rep makes an estimate, which rolls into managers, who eventually present their forecast to a leader. Some managers allow reps to make an estimate based on dollars– without specifying which deals will close, while others force attention on individual opportunities. Either way, by the time the forecast makes it to company leaders it’s an amalgamation of so much data that it’s difficult to determine what’s important. No wonder most forecasts are inaccurate!

With AI now putting the sales forecasting process under a microscope, leaders are discovering that only a small percentage of closed deals were predicted accurately in advance.

Why do reps over-forecast? When are reps most likely to under-forecast?

Inaccuracies in the forecasting process more often than not are a result of human behavior. Reps tend to over-forecast because they believe their deals will close or are unable to admit when a deal has gone south. On the other hand, if a rep has already hit their quota for the period they might try to push remaining deals into a future period by under forecasting.

The ideal forecasting process eliminates emotion and personal agendas to get to the ‘closability essence’ of the pipeline.

How does over- and under-forecasting impact sales leadership? According to data from our recent study, which analyzed over a quarter million sales opportunities, only 28.1% of closed deals (closed won or closed lost) were predicted accurately by sales teams 90 days out. This is worse than flipping a coin and guessing the outcome of the toss.

Inaccurate predictions have a big impact on sales leaders in how they manage their reps. Over-forecasted deals draw the attention of sales leaders when they shouldn’t, and under-forecasted deals never get the attention they deserve. In a world where emotion and personal agendas drive the forecast, this is always the problem. But sales leaders are stuck in the middle–they can either trust their reps and focus their attention on the wrong deals, or they can try to understand every single deal and make their own judgement– something almost no sales leader has time to do.

Why do businesses need to pay attention to how drastically they over pitch?

Businesses are run, scaled and grown on their leaders’ ability to accurately forecast revenue each quarter. The problem is even when sales leaders correctly predict which deals will close, the size of each opportunity varies so widely it becomes nearly impossible to make an accurate forecast.

Today, companies have a lot to gain by replacing highly inaccurate rep-based sales forecasts with an approach that uses predictive analytics to take the emotion out and generate unbiased projections.

How can AI help solve this problem?

AI was built to solve forecasting problems. In non-sales applications, machine learning algorithms look at all available data about historical circumstances and correlate those with outcomes. The algorithms are then tuned to predict outcomes and tested against history. The same can be done for sales forecasting and is being done today for some of the most sophisticated sales organizations.

For sales forecasting, machine learning uses historical opportunities, the observable characteristics of those opportunities, and their eventual outcomes to determine which factors influence an opportunity to be won vs. lost:

Is the type of prospect we usually win?
Is the opportunity configured such that we’re likely to win (product, price,etc.)?
Are we doing the right activities?
Is the deal progressing such that it is likely to close in this quarter?

AI takes all personal bias out. It provides insight to managers and leaders on exactly which opportunities are likely to close. Now a leader can scan an opportunity-based forecast and quickly determine whether a rep is over- or under-forecasting. No matter what process a leader uses to roll up their forecast, this is a good sanity check. AI can focus our attention in the right spots to help create the most accurate sales forecast– changing the sales forecasting process forever.