In a World of AI Co-Pilots, Who’s Ready to Fly the CX Plane?

In case you have not heard, artificial intelligence has become a full-fledged “thing,” yanked off the experimentation floor and embedded in almost every single system and solution you use today. Across the spectrum of customer experience (CX) we have seen AI co-pilots positioned to guide, aid, and take the mundane off of sellers’, agents’, and marketers’ plates.

In revenue optimization, tools are available to aid sellers engage with and communicate with buyers at the right time and right space. In service, co-pilots help agents in a very similar way, up-leveling recommendations for responses, follow-up, and proactive engagement opportunities. AI co-pilots are also supporting supervisors, analyzing call volume, sentiment, quality, and outcomes to help best coach, plan, and develop winning game plans for optimization of time, resources, and revenue capture. Marketing is also in on the co-pilot party, using AI-powered assistance in crafting emails, personalizing content across all channels, timing and orchestrating journeys, and taking audience segmentation to new finely tuned heights.

Co-pilots are popping up everywhere, learning and expanding their recommendations, honing their capacity to aid and assist with every new data source, gaining a better understanding of how we work and how our customers are reacting and responding.

There are also signs that consumers are warming to the idea of AI (especially generative AI) becoming a part of their everyday lives. In EY’s Future Consumer Index research, there was a sense that one doesn’t necessarily have to understand AI to enjoy using it. Only 22 percent of U.S. consumers felt they had a good understanding of the technology but 50 percent felt comfortable using AI to improve purchase experiences.

This comfort in using AI tools is also reflected when you dive deep into specific applications of AI. In the banking space, the financial services technology vendor FIS released a survey noting that 58 percent of U.S. consumers say they would be comfortable with their bank using genAI applications. Other studies from MITRE-Harris Poll back this sentiment up in broad general industry applications as well, noting that consumers, by and large, are more excited than concerned about AI. In fact, 51 percent of men, 57 percent of Gen Z, and 62 percent of Millennials are in what we might call the “hyped-up” camp, leaving us Gen X women to hold down the anxious and worried brigade.

Perhaps that’s why the question of pilot readiness has entered the picture. After all, if these AI tools, bots, and workflows are being billed as the ultimate co-pilots for the work and workers across CX, at some point we need to recognize the pilots (more colloquially known as “people”) and the planes we are all flying. Perhaps the time is now for the pilots of the CX plane to update and refresh that preflight checklist, ensuring that we understand the makeup of the craft we oversee maneuvering from point A to B so that we can more effectively rely upon, and grow to trust and partner with, the co-pilots built on our behalf.

Here are just two items this age of the AI co-pilot is giving us license to take advantage of.

Training and upskilling: Regardless of the stage or phase of one’s career, adding some new skills and tricks specific to the AI and AI applications that have been tailored to your functional group is the new imperative. A quick scan of the e-learning platform Udemy shows that everything from AI (with 3 million-plus learners enrolled in courses) and machine learning (8 million-plus learners) to ChatGPT (2 million-plus learners) and deep learning (2 million-plus learners) is a hot course to join. While learners may be in these courses in the hopes of launching a new career in a burgeoning space, others are there to keep up with the enterprise Joneses.

From understanding why a prompt returns better results to finding new processes to be automated or offloaded to a highly qualified bot/co-pilot, upskilling in AI could become a mandatory round of continuing education for any team member in CX. In the same way airplane pilots have an established cadence of retraining and upskilling to stay ready and modern, those of us working with AI co-pilots will likely need to follow a similar regimen, brushing up on prompt skills, getting insights into underlying models, or just taking a few moments to play with emerging tools to gain added comfort in knowing a bit more about how co-pilots work and how they learn.

Oversight and orchestration: For managers of teams being introduced to AI co-pilots, think of this has having two different, yet connected, workforces that you must motivate, manage, and celebrate. Manage the bots? Performance reviews for the IVAs? What next—a performance improvement plan for the lowest performing sales co-pilot? If that’s what it takes, yes. At least in the short term.

Successful co-pilots will be those that are used and those that are valued. But there will also come a time when these investments will need to deliver, making an impact on both bottom- and top-line performance. Start itemizing where you, as the manager of this new league of co-pilots and assistants, believe you can reasonably expect these points of impact. How well are the pilot and co-pilot working together? Is data quality or a lack of data impacting the capacity to deliver value? It might sound trite and overly simplistic to read these questions in black and white…but the reality is that there are mountains of data, recommendations, blinking alerts, and flashing screens our teams are about to bear witness to as co-pilots native to the tools of engagement start to compete and potentially clash with the alerts and flashing screens of those now available via APIs, marketplaces, integrations, and partnerships.

Having a sense today of reasonable expectations for this new human-AI wonder team will help you quickly and firmly make decisions around training, upskilling, or turnover when a co-pilot just isn’t making the grade.

The AI co-pilots we welcome into our stacks and racks today are just that—co-pilots, set to focus models on specific tasks. But if you allow these models to multiply, collide, and run amok with little business oversight or intention, then those instances of unintentionally negative moments of opportunity—as an example, the days of sending an automated cart-abandonment email reminding a customer about the red shoes they looked at last week, without the necessary orchestration or guardrails to identify that the customer not only bought the red shoes, but has already returned them—could take on a whole new level of madness. This time, these crossed wires and random acts of impersonalization will bring down our pilots right alongside the AI bots and co-pilots. 

Liz Miller is vice president and principal analyst at Constellation Research, covering the broad landscape of customer experience strategy and technologies.