Calculating Deal Close Probability
Sales forecasting tools and methodologies do a wonderful job in making forecasting look easier than it is. It seems like all you need to is put your deals in stages and then multiply each deal by the close probability of the stage – 10%, 25%, 50%, etc – and you’ll have your forecast number.
What is rarely talked about, however, is how those magical numbers of close probability for the deal stage got there in the first place. Interestingly, if you look at many organizations’ stage probabilities, they look suspiciously perfect, with nice round numbers like “10%” and “50%”. In fact, when pressed, they will admit that they took a wild guess when they first put the system in place, and have not come back to revisit the numbers since.
Given that these close probabilities are the foundation of the pipeline forecast, it seems worrying that they are essentially fabricated numbers.
Looking Back At History to Find Probability
When articles do look at more systematic ways to calculate deal probability they usually start from looking at historical data. It’s a fine way to start. If you take all the deals that have been in a given stage, and then look at the percentage of those that have been won, you have a good sense of the probability that you will win any given deal in that funnel stage.
However this approach relies on the deals that have been in that stage historically being correctly assigned. If sales reps are allowed to self assess deals, they have a tendency to carefully “massage” the data to move deals around so their pipeline looks how they want it to look. This makes the historical view somewhat inaccurate.
Deal Stage Correction with Objective Data
Before looking back at historical data to calculate close likelihood, it’s worth doing a “clean-up” to remove the deals that should never have been in that stage. For this, you will need to have an objective view of what it should mean to be in each stage.
The best approach to this “objective” view of a stage is to look at relationships to understand how high level, and how broadly your team was able to get engaged at the account. If they only had one weak relationship, the deal could not have moved beyond “Prospecting”, if a few more relationships were built, “Discovery” could be a valid stage. “Business Case” would require at least a medium strength relationship with an executive with a Vice President or higher title, and “Negotiations” would imply that legal or finance roles were involved.
Looking at past deals through this lens allows you to clean your data set to remove deals that should never have been marked as being in that stage. If the same objective criteria are used for current deals, you can be confident that the percentage you calculate based on historical deals is quite accurate for current deals.
Sales Forecasting Relies on Deal Probability
An accurate sales forecast relies fundamentally on an accurate calculation of the probability that a deal will close. To get this, we need to calculate based on deals that have been in each stage historically. This calculation is futile if there is not an accurate sense of which deals correctly should have been in each stage, and to do that we need to understand who was involved in each deal, what their seniority level was, and what their functional role was.
With accurate data on historical selling patterns and buyer relationships, we can see deal stages accurately. With accurate history of deal stages, we can calculate a true close probability for each pipeline stage, and with accurate forecast probabilities we can get closer to an accurate sales forecast.