Keys to personalizing AI for sales

For businesses, there are only two reasons to deploy artificial intelligence (AI): to make money outright or to lower the cost of making money in the traditional fashion. To make money, there is no other option but to unleash AI on the sales process itself.

Since sales is such an integral function within the business model, companies should understand how AI works as thoroughly as possible, if only to recognize what it can and cannot do to move products or services and to increase revenues.

Tech leaders

Perhaps not surprisingly, the tech industry is ahead when it comes to leveraging AI for sales. Intel, for one, is already five years deep into a program to supplement both sales and marketing with advanced intelligent analytics. The objective is twofold, says Venture Beat’s Kyle Wiggers: to gain actionable intelligence into customer buying patterns and other trends and to guide company resources in the right direction. Two key applications are Sales Assist and Autonomous Sales — the former broadens managers’ customer interactions so they can oversee more accounts, while the latter drives automated sales through email, web ads, newsletters, and other forms of outreach. Sales Assist already counts more than 1,500 users at Intel, while Autonomous Sales delivers more than 30,000 emails per year, which generates a first-time purchase conversion rate of about 16%.

Divining buyer intent is AI’s superpower when it comes to sales, says David McFarlane, a board member of the MIT Enterprise Forum. Unlike traditional predictive analytics, AI can quickly assess thousands of website visits, data pathways, and hundreds of other digital footprints. It can then compare them to millions of profiles and previous interactions. In some cases, this is pushing success rates from the single digits to more than 90%.

The goal here is not to turn low-probability buyers into high-probability ones, but to quickly identify the high-probability buyers and engage them with the right tools, like a messaging platform or a search engine. It can do this even if the buyer remains completely anonymous. At the same time, low-probability users can be served with tools appropriate to their experience, like a chatbot trained for either proactive or reactive service.

All of this is leading to the era of AI-assisted sales, and not a moment too soon, says Allego’s Ginna Hall. Like everything else, sales are starting to scale up to digital proportions, which would require an army of human reps to keep up with the load. By equipping sales professionals with AI tools, organizations will be able to target all of their resources in the appropriate manner and rapidly scale up their responses as needed.

A human touch

In this hybrid world, AI can actually provide a more personalized sales experience for customers. Most sales staff, after all, are trained on a one-size-fits-all model that tends to lump buyers into various groups. With AI’s broad access to data, however, it can be trained to tap into the unique requirements of the individual at far greater scale than a human sales force. Perhaps, this is why two-thirds of companies report increased revenue from their AI-assisted teams, according to McKinsey.

As with any technology, however, there are right ways and wrong ways to implement AI into sales. Perhaps the worst way is to throw AI at a problem without giving it access to the right data. Diane Braybeal, of the Marketing AI Institute, points out that automating data collection, filtering, conditioning, structuring, and other tasks is key to successful outcomes. With good data, AI can then be trusted to deliver more accurate lead-scoring, more personalized experiences, higher levels of upselling, reduced churn, and generally better sales management by improving sales targets and quotas, optimizing territory, and predicting the impact of staff and product changes.

If there is one overarching rule in deploying AI into sales, it should be to avoid giving customers the impression that they are being manipulated by robotic processes. Consumers are already getting wise to the bombardment of ads tied to their most recent search queries or even private conversations with loved ones. For AI to succeed, it must seem as normal and natural as possible, or at the very least, not forced or opportunistic.

People still want to be treated like people, even if they know they are dealing with a robot.