In the Age of AI, Training Is More Critical Than Ever

Erik Charles

When sales professionals look at artificial intelligence (AI), what do they see? If they’re technology-averse, they might see a threat to their jobs. If they are big fans of technology, they might see a silver bullet that instantly boosts their results.

Both of these visions are as extreme as they are wrong. AI is not going to make salespeople obsolete, nor is it going to turn bad salespeople into great closers. What will it do? It will become every salesperson’s most efficient assistant. And, at that point, the winners and losers will be separated by how well they combine their sales talents with skills for using AI.

Before that happens, though, the AI has to be trained —  that is, subjected to a set of data that allows it to make correlations that are then compared to predetermined “correct” and “incorrect” answers. So, the first thing an artificial intelligence application needs is data. For example, sales applications should draw on a focused set of data — from CRM, sales performance management solutions, sales enablement, CPQ and other sales-related applications.

Additional data sources

One of the risks in much of the current work being done in AI (and machine learning) is only looking at internal datasets. This is risky, as it captures how your firm might perform, but that does not mean that your firm is right. Bringing additional data sources to the system helps ensure that you do not simply make your existing “worst” practices more efficient. This could include data from relevant industry organizations, government economic indicators and various inflationary measures for different geographies.

The first step toward sales AI success will be taken not by salespeople, but rather by sales ops and the IT pros determining the right data to train the system. They must then develop, through machine learning, a system that delivers the right answer nearly all of the time.

Once that’s accomplished, AI will arm salespeople with content, upselling and cross-selling suggestions, and help with decisions virtually instantaneously. It will help salespeople deal with a spiraling complexity of products, and with increasingly large catalogs, ever-more complex compensation plans that include SPIFs and bonuses, and financing options to help make bigger deals possible.

Selling skills still matter

The thinking is that AI will turn B-players into A-players, and A-players into unprecedented selling machines. But this makes a major assumption: that salespeople will know how to use the new bounty of knowledge at their fingertips. If mere access to data was the key to sales results, the widespread use of CRM should have had a “rising tide lifts all boats” effect on sales. But, according to data from CSO Insights and Buyer Zone, in 2012, 56 percent of all businesses used CRM and 63 percent of salespeople achieved quota. Compare that to 2016, when CRM penetration surged to 74 percent but goal attainment dropped to 55.8 percent, and it’s clear that data by itself is not the key to sales success.

AI will not supplant selling skills; A-players will likely be better at using assistive intelligence than B-players. This creates a new challenge for sales management: how do you train your sales team — including the B- and C-players — to use your version of AI to maximum effect. This is unexplored territory — and again, sales management, sales ops and IT must work together to develop an effective strategy to teach people how to combine information delivered by AI with their own skills in order to influence deals.

Noted analyst Esteban Kolsky wrote in a recent blog post that AI will never match human intelligence because it lacks the three I’s: intuition, imagination and innovation. These are the skills that humans possess that are impossible (at least, for now) to build into an AI system —  and they’re the skills that will determine how effective salespeople are at using AI to drive sales results.

As counterintuitive as it may seem, the secrets to sales AI success is entirely dependent on training: training of the AI itself and training of the salespeople who must use it. Without those two activities, AI is little more than a novelty that will distract your sales force from its real objectives.

Erik Charles is the vice president of strategic marketing with Xactly, which provides cloud-based enterprise and software services. He can be reached at Follow him on Twitter @erikchaz or connect with him on LinkedIn at