The best bit about knowing when you’re going to make a sale is when it actually happens and you’re fully prepared to bathe in the glory. Sadly, as humans, we’re pretty terrible at judging this. In its 2016 sales enablement optimisation report, CSO Insights put win-rates for forecasted sales at less than half (46.2% to be precise). Alarmingly, that’s actually eight percentage points less than a year earlier. At that rate, no businesses will be closing any forecasted sales within six years!
If stats aren’t enough to convince you of our poor powers for prediction, let me take you back to 2013, when Alan Greenspan, as a guest on the Daily Show, told Jon Stewart: “We really can’t forecast all that well and yet we pretend that we can, but we really can’t.” That’s Alan Greenspan. Former chairman of the US Federal Reserve! If that lot can’t get it right, what hope is there for the rest of us?
Well, as it happens, there is cause for hope. What if it wasn’t people doing the forecasting? What if it was an emotionless machine unburdened by the shortcomings of humanity and dealing only in cold, hard data? What if you brought together predictive analytics with sales forecasting to create predictive analytics sales forecasting – the use of machine learning to crunch your data and make forecasts about sales probabilities? You’d forecast more accurately. Much more accurately, in fact.
Don’t take our word for it, though. When guest authoring an article for Entrepreneur Magazine, Justin Shriber, the then vice president of products at predictive sales and marketing outfit C9, placed the accuracy of predictive analytics for sales forecasting at 82%, which compares very favourably to either the 54% or 46% figures that CSO Insights has variously attributed to human sales forecasting. If you’re wondering how that worked out for Shriber, C9 was acquired a matter of months after that article was published by Inside Sales, which was recently named on the Forbes Cloud 100 list for the second year running.
“How can this be?” you ask. Well, predictive analytics platforms can crunch a far greater number of variables and can pick out patterns and trends that are both more pronounced and less clear than people can. Using our own platform to illustrate the point, we’re able to take a company’s sales and pipeline data, information about its prospects and even unstructured data like emails and social media interactions to develop a predictive model for the business. By identifying correlations within the data, the platform can not only spew out accurate predictions about future deals, but becomes more accurate the more data it is fed.
What all this means, in short, is that it’s possible to forecast sales with a high degree of accuracy using data you already have, as well as combining it with data from other sources. In addition, though, as a by-product of the predictive process, you can identify the probability of getting individual prospects over the line and apportion your sales team accordingly.
You can get opportunities that would have completely passed you by as a result of algorithms spotting things that you can’t and you can even automate large parts of the sales process, like assigning leads to salespeople and sending initial emails, so as to free up your sales people to spend time closing and celebrating the deals they’ve closed. Finally, you’ll be in a better position from which to plan and optimise working capital, inventories, manufacturing, supply chains and logistics. I’m sure you’ll agree, it’s not just which sales you’ll make that’s nice to know about in advance…