CX leaders find they are in the same GenAI boat—and it’s not moving nearly fast enough

New research across the Global 2000 shows that 10% of tech budgets are being dedicated to GenAI. However, the customer experience function is not reaping the benefits.

Ambitious customer experience leaders must do a better job of educating themselves and their teams about GenAI’s potential, lose fear, and embrace emerging opportunities now. Otherwise, they will be left in the wake of organizations that have already set sail for the future.

More than 500 business leaders surveyed by HFS (see Exhibit 1) identify improved customer engagement and satisfaction as both a quick win and a top-rated benefit of GenAI. Yet a roundtable of leading CX executives, hosted by HFS and Firstsource at London’s Dorchester Hotel, exposed the reality that many leaders in the function feel their ability to embrace generative customer experience remains a distant fantasy. The common consensus was relief because while they all thought they were lagging in capitalizing on the benefits of GenAI—at least the leaders had discovered they were all in the same boat.

Exhibit 1: Leaders from across the enterprise expect GenAI to deliver quick wins in CX (improved customer engagement and satisfaction)—but many CX leaders remain hesitant

N=550 business leaders identified top business benefits from GenAI in the enterprise and when they anticipated realizing full benefits.
Source: HFS data

What’s holding CX leaders back?

CX leaders spoke candidly about their concerns based on their experience evaluating, piloting, or edging use cases toward production.

Use cases their firms are working on include:

Customer interaction:

• Customer chatbot and knowledge base search
• Customer service personalized automation
• Translation, text, and voice
• Voice of customer analysis

Employee support:

• Creative design
• Optimize email/Outlook usage
• Ticket assist
• Frontline staff solutions
• Company branded copilots

Software development & IT:

• Release management
• Debug builds
• Quality assurance FAQ creation
• Network monitoring capabilities

But despite this apparent flurry of activity, many leaders remain unconvinced that they, their teams, and their customers are ready for GenAI.

Among the concerns voiced by roundtable attendees was the discomfort some employees felt when asked to work with GenAI. They also noted the challenge of serving a wide range of customers—many of whom have not yet come to terms with speaking with a machine.

We are still in a world where many customers won’t fill in online forms, want to talk to a person, or don’t trust that bots can handle their complex inquiries.

Who is training whom? Are humans training bots or bots training humans?

One leader said she felt that the form of “natural language” we use to obtain responses from generative AI impacts our expectations and is defining how we act toward the bots, raising the question, “Who is training whom?” Are the bots training us to interact with them in the best way for the bot rather than vice versa?

Our group of executives also expressed concerns that bots could not generate the empathy of a human operator. Recent developments in GenAI (ChatGPT-4, for example) do a reasonable job of mimicking empathy. Still, the group felt there always will be occasions when customers want the human touch. CX execs see bots as companions to human operators. When we deploy bots to respond to customers, they should work “with a human hand on their shoulder.”

CX leaders’ concerns are acting as GenAI roadblocks

Concerns about ethics and the need for responsible, non-biased decision-making remain at the top of CX leaders’ minds, too, just as they constrain the pace of adoption across many other enterprise functions. Where we deploy bots in customer-facing applications, our fear of unwelcome hallucinations rises, increasing demand for robust and transparent governance.

Exhibit 2: CX Leaders GenAI’s value increasing in the next two years

Source: HFS Research 2024
Sample: 22 UK roundtable delegates

The CX leaders at our roundtable are—on average—expecting to find GenAI increasingly helpful over the next two years. However, they also think the benefit it will deliver will be more personal than beneficial for their enterprises.

This skew to personal productivity indicates how early in their journey our CX leadership group may be—openly acknowledged by those around the table. HFS Data gathered from enterprises at the leading edge of GenAI implementation identifies an expected 10-20% positive impact on all business levers in the next two years.

That CX leaders see personal benefits coming faster than enterprise benefits at this stage suggests that their concerns are blocking decision-making about GenAI. They must overcome these blockers to ensure they are not left behind as more decisive competitors enjoy 10-20% improvements in business benefits.

What CX leaders must do to move faster on GenAI

Most CX leaders are on a steep learning curve regarding GenAI. Leaders must immerse themselves in the capabilities of GenAI by using examples of it in their work every day. Regular personal use of the technology leads to familiarity with its capabilities. Regular use of GenAI will help them understand its limits and gain empathy for what is causing customer and employee fear.

They should work with peer groups, their service partners, and—yes, research analyst firms such as HFS—to ensure they are learning from each other and getting good, expert guidance. Many service providers are further along in their GenAI journeys than the enterprises they seek to serve. Like other business leaders, CX leaders can take advantage of the lessons service providers have learned in developing responsible AI and robust guardrails and governance, which will remove many of the concerns CX leaders expressed at the roundtable.

The traditional concept of CX is changing exponentially. We’re actively listening to our customers’ unique business challenges and really working to define, implement, and deliver the business benefits of tailored GenAI solutions. Our goal is to help them reimagine and redefine what world-class customer experience could and should look like.

– Rajiv Malhotra, President EMEA at Firstsource

Small Language Models may prove a faster, cheaper solution for customer service chatbots

The relentless demand for instant ROI may hold back some organizations. While returns are important, the group heard of one case in which a multinational supermarket chain made investments with the sole initial intent of learning the power of large language models (LLMs).

Once the firm learned about LLM capability, it could more effectively target investment at the correct business problems, applying LLMs where they could have the most impact.

When scaling, many organizations are hitting new challenges—overcoming latency (which can be as long as 12 seconds in some LLMs) and rising costs. Small Language Models (SLMs) are emerging as a solution that offers the adaptability and responsiveness that real-time applications—such as customer service chatbots—demand. The smaller model size means firms can run models on local machines at lower cost and with greater control over potential data leakage.

The Bottom Line: CX leaders must use the ecosystem of their peers and suppliers to take their rightful place at the front of the pack

CX Leaders should be at the forefront of GenAI adoption. Many are being held back by concerns that others are already solving. Leaders should act now to get personal with GenAI daily, connect with external expertise, and partner with peers to accelerate their journey—or risk missing out on benefits their faster-moving rivals will reap.