Harnessing AI in Modern Customer Engagement Centers

Customer service satisfaction in the United Kingdom is plummeting. Forget robots taking all the jobs; the culprit might be the very technology designed to streamline interactions. A relentless focus on digital-first strategies, prioritizing automation over human support, seems to be backfiring. Customers are frustrated by complex menus and missing phone numbers, while poorly trained agents struggle to handle increasingly intricate interactions.

I'm not a fan of the trend to change well-established job titles and department names. These changes often seem unnecessary and can even make it harder to understand what the role or department actually does. Consider the term customer engagement center. Phrases like customer contact or customer service center are clear and concise, accurately reflecting the department's purpose. Personally, I value service and prefer the latter label.

Changing it to a customer engagement center is already re-positioning what a consumer can expect. It is almost setting out to say that our function isn't to serve you, it's to deal with you, almost as if you are part of the problem not the reason the business exists. It is a slippery slope, as is the often-inappropriate use of artificial intelligence (AI) in customer service centers.

Technology (certainly AI) undoubtedly has a place in the customer service center of the future, but, like many things in life, it is a balance and is often fraught with unintended (often negative) consequences.

The Institute of Customer Service's U.K. Customer Satisfaction Index tells us that customer service standards have continued to fall to an all-time record low. There are a number of factors that are driving this sustained and continual slump. One is what many firms have phrased digital-first, a strategy to deploy technology to enable customers to self-serve and be served by or through technology, thereby replacing a human being.

It has the clumsy footprints of distant accountants all over it. Deploying technology enables us to reduce headcount, and that reduces the cost to serve. Hard to argue with the numbers in the spreadsheet. The trouble is that the accountants, in their enthusiasm to bank cost reductions, forgot to ask if customers wanted digital-first. It turns out that the vast majority of customers actually want digital-on-demand, which is very different than forcing customers down a certain route whether they want to or not. Concealing inbound telephone numbers and making it painful for customers to engage via the phone doesn't feel like a sustainable strategy for companies that care about the service they provide their customers. At a point in history when it has never been easier to switch suppliers, such a strategy might backfire painfully.

One unintended consequence of digital-first has been that much of the easy traffic has been removed from the agent's workflow, meaning that almost every interaction is more complex, demanding a very high level of in-role competence from the agent. This isn't a problem per se, provided that the employer has taken account of the requirement for an elevated agent skill set and the learning consequences of hybrid and working from home, which is certainly the reality for most customer service centers right now.

The learning challenge of hybrid working is that as humans we develop about 70 percent of our in-role competence from peer-to-peer learning in the workplace. Inconveniently, not through low-cost, online tutorials and computer-based training (CBT). If as an agent you are working three days at home and two days in the office, then you will be deprived of at least 60 percent of the peer-to-peer learning opportunities you would have had if you were 100 percent office-based. And this assumes the people from whom you would have learned are in the office on the days that you are. Which, statistically, they won't always be.

I have no problem with hybrid working, I have personally done so for more than a decade, but then my role is very different from a customer service agent. We know from thousands of interactions with employers across Europe, the United States, and Asia, that few firms have materially changed how they train and support their employees post-COVID-19 migration to work from home and hybrid working. As a result, agent competency levels are falling well short of what is required to support the now continual stream of more complex customer interactions.

These aren't my personal conclusions either. The UKCSI report found that poor staff competence is one of the top three drivers contributing to declining customer satisfaction. Only quality or reliability of goods/services and late delivery or slow service were ranked higher by U.K. customers as the leading two drivers.

With inadequate training and competency development programs, coupled with the growing complexity of customer interactions, it is not a coincidence that some sectors are reporting record levels of new employee attrition during the first 180 days of employment. Agents are leaving because they don't have the skills, knowledge, support, and confidence they need to do the job. It is simply not enjoyable and too stressful, and they can earn the same money or more in the gig economy.

If employers don't act to fix the problem, then the next issue will be an inability to recruit and retain agents. The solution to which will be to raise wages to the point where employees will put up with the stress , further infuriating the accountants!

There are excellent AI-driven technology solutions that will drive mutual benefit to the customer and service provider. They are unlikely to be universal solutions and will inevitably operate alongside human agents. And smart companies will recognize that digital-first is a broken model, and customers want to choose how they interact with their suppliers. While this might irritate and frustrate the accountants, because obviously it would be much easier and cheaper if the customer just did it all themselves, it does mean that employers need to upskill their agents who will be operating in this digital-on-demand environment.

The failure to adequately upskill and right-skill agents is perhaps the biggest unintended consequence of the digital-first and AI revolution sweeping the customer service center world. Failure to train agents appropriately drives costs and dissatisfaction into the business. But training isn't the required outcome; competent employees are, and achieving this, especially in a hybrid environment, has proven difficult for most firms.

One consequence is that agents are appropriately competent to perform their role. Another is that employer investments in other AI tech are more likely to succeed and deliver the forecasted cost savings, as the human agents better understand how they integrate with that technology.

The lesson we can learn is that most customers value good service and humans will form a component part of that service provision for many more years. To optimize the deployment of a mix of technology and human agents, firms must invest in right-skilling their workforces. Hoping the human elements just work it out on the job is a poor strategy akin to hiding your phone number. Ironically, AI exists that not only solves the problem, but also increases the value a human agent delivers the employer and makes the role far more interesting and rewarding.


Adrian Harvey is co-founder and CEO of Elephants Don't Forget.