CITE World: Predictive analytics: A look into the future of sales

July 16, 2014-

By Erika Morphy

Another day, another Salesforce acquisition. Last week the CRM giant acquired RelateIQ, a relationship intelligence platform. This is more, though, than just a story about a giant software company acquiring a missing piece of functionality it wants to add to its mix of offerings, although it is, of course, that too.

Rather, the real story here is about the technology that Salesforce is acquiring – predictive intelligence – and where it is presumably being added: the sales process.

Actually, the industry has been chasing after this ideal for a while, with mixed results. Now, though, thanks to a mix of advanced algorithms, self-learning technologies and, in some cases, wide latitude given to human input – yes, that ‘gut feeling’ that sales reps will swear by – CRM may be ready to add predictive analytics to the sales process.

Simply put, predictive intelligence has been missing from sales, says Clari CEO Andy Byrne. The RelateIQ acquisition, though, is only tackling the “relationships” piece of the sales puzzle, he says.

“To close more deals, reps need intelligence on deal progress, and actionable information on how to close deals. In turn, managers need better information for their forecasting,” Byrne says.

When this happens – or as it happens – it will look something like this: the rep has various deals in various stages of the pipeline. He or she diligently enters the data necessary to keep the pipeline current; in other instances, the system updates itself with these data points automatically when able. The rep, and his or her manager, is able to keep track of these deals and predict whether or not they will close in real time. The system will be able to say that ‘Deal XYZ has a 49% chance of closing by quarter’s end because you sent the proposal and followed through afterward and there are signs that the company is getting ready to expand into a new market and it will need the product that you are selling.’ Or something to that effect.

As with anything software related, it is not that straightforward, of course. Some of the apps emphasize the data mining and corporate intelligence aspects; other focus on self-learning and adjusting predictions as more deals enter the pipeline. Perhaps the best description of the predictive sales category comes from a blog post by Forrester analyst Mike Gualtieri. Essentially, he says, a good predictive analytics app will:

• Learn who the customer really is
• Detect the customer’s intent in the moment
• Morph functionality and content to match the intent
• Optimize for the device or channel

The Human Factor

He might have added, though, that a good app would incorporate the sales rep’s insights into the mix. Consider this scenario: a higher up in a company attends a demo of a product and appears to be enthusiastic about what it can accomplish for the firm. Normally that would be a plus in the sales rep’s mind. Or maybe not. Much depends on the particular industry, and perhaps buy-in from that particular branch of the company or even that particular executive isn’t as necessary as it would seem.A sales rep might just well get that gut feeling this person isn’t important—maybe no one else at the demo was paying that close attention to his questions—and decide to focus on the a lesser-titled team leader in the back of the room who seems to stop conversations whenever he has a question.

The new tools entering the market account for and respect these instincts, says Neil Biehn, data scientist and vice president of science and research at PROS. In general, he says, this is a tricky business.

“We’ve all learned that uncorrelated events can trip people, making them think a certain action made a prospect go away but it may not have been related at all,” he said. The right tool, he said, supports the sales process striking the right balance between providing data or information and taking into account a sales rep’s experience.

Some of the apps in this space focus on a specific portion of the sales process, such as PROS’ Cameleon CPQ. The company recently demoed the Configure-Price-Quote product on the new Salesforce1 platform, showing how reps can arrive at faster quotes and in general improve their sales effectiveness using the system. PROS also pointed to an Aberdeen Research report that shows that companies that employ CPQ solutions enjoy a 105% larger average deal size, 27% shorter sales cycles and 19% higher lead-conversion rates.

Other apps focus on determining how likely it is a deal will close. Clari launched a mobile sales productivity platform earlier this year that uses a number of building blocks, including the rep’s data and contacts, commercial databases and internal applications, a database of previous deals, and real time intelligence, to come up with a finding or projection that is, as marketers like to say, actionable. The system will compare current deals with similar transactions from the past and then calculate whether it will close — and what actions a rep might take to nudge that projection upward.

Salesfusion does something similar with a marketing automation platform it rolled out earlier this year, which was enhanced with LoopFuse’s predictive lead scoring capabilities. The system evaluates sales and tells the reps which will close and which won’t, based on what actions they are taking or not taking. It makes these decisions based on its previous knowledge of past deals and the industry in question.

Because it is a self-learning app, the system improves and updates itself every time a new deal closes and more information is added to the mix. The sales person also enhances the self-learning aspect of the system by weighing in on the leads the system has rated as hot or not, so to speak. By giving a particular decision the thumbs up or thumbs down the system learns even more about what work and what doesn’t.