AI brings promise and peril to customer relations management

Artificial intelligence is proving to be most valuable when it is applied to rote, predictable functions. At first blush, this may not sound like an ideal fit for customer relations management (CRM), but keeping customers happy requires a lot of tedious work.

In today’s increasingly digital world, CRM lives and dies by data — not just the amount of data gathered but the quality of that data, which can only be achieved by cutting-edge analysis and interpretation. But today’s volumes are simply too much for human analysts to cope with (at least, in a timely fashion), so CRM platforms of all stripes are starting to incorporate AI to handle the load.

Bad data, bad decisions

Faulty data management is a key factor in poor customer relationship outcomes, with more than 85% of sales agents citing it as the cause of embarrassing mistakes, according to marketing analysis firm MarTech Series. Upwards of 2.5 quintillion bytes of data are being created every day, 90% of it unstructured, so the mere task of putting all of this data into context is functionally impossible without AI. And AI can be integrated directly into CRM workflows to handle the tedious tasks most people don’t want to do anyway — and usually can’t do without introducing numerous errors. At the same time, AI can be trained to communicate directly with customers, either by text or voice, to address simple questions or easily solvable complaints.

As counterintuitive as it may seem, AI is likely to produce a more personalized approach to CRM than is currently possible. AI can assemble and assess a customer’s digital history — including purchases, emails, and other events — to determine their needs and temperament much faster and more thoroughly than a human representative could, Anzhelika Danielkievich recently wrote on the Keen Ethics blog. This helps resolve problems in a timely manner, and with a higher degree of satisfaction, but it also enables a more accurate representation of brand sentiment to further hone marketing and communication efforts.

In addition, AI can do wonders for much of the behind-the-scenes work of CRM, such as lead scoring, cross-selling, price optimization, and sales forecasting. This information can then be used to improve business strategies, right down to targeted advice to sales reps to guide them through each stage of the sales pipeline.

A friendly voice

The ultimate goal is to provide better customer service, according to software developer Nahla Davies. At the moment, one of the chief complaints aimed at companies large and small is the long wait times at call centers and in email replies. A properly trained AI-driven CRM platform will be able to handle most common queries with little to no delay, sending the more complicated requests to service reps, who should have greater availability. AI will also be able to more effectively communicate with customers over the web, social media, and mobile platforms.

AI can also help people interact with enterprise services in a more streamlined and secure fashion. Davies notes that mobile banking is already pushing software that allows customers to take complete control of their finances, with AI programs constantly monitoring for threats and then pushing out the appropriate updates to security tools like encryption and two-factor authentication.

As with any software, it’s important to note that all AI-based CRM platforms are not created equal, nor are they immune to vendors’ tendency to overpromise and underdeliver. For instance, all the talk about almost humanlike interaction between customers and service bots tends to skip the fact that this level of technology is still a few generations away. Right now, AI bots are being purpose-built for specific use cases, such as data entry and task scheduling.

On the other hand, AI is giving a major and immediate boost to functions like predictive analytics for everything from trend and market forecasting to driving inefficiencies out of supply chains. In the end, enterprise executives should take a hard look at what AI can do best and target it at specific areas where its efficacy can be measured against established metrics.

Perhaps the most important thing to keep in mind when augmenting a CRM with AI is that the technology should be a conduit to successful outcomes, not a barrier. Many people will gladly engage with AI if it provides a quick, simple resolution to their problem. But frustration will mount if the problem is not solved and they can’t get past AI to a human representative.

Likewise, customers will likely be unhappy if they are talking or texting with what they think is human but turns out to be a bot. The use of AI should be made clear up front, and it must be used at the customer’s discretion.