Customer experience is facing a reckoning. After years of investment in digital tools and AI, Forrester’s latest US Customer Experience Index shows CX quality is at its lowest point since tracking began. AI-driven chatbots, once heralded as the future, have played no small part in the decline.
CXM spoke with Joana de Quintanilha, VP principal analyst for customer experience at Forrester Research, to unpack the early missteps, the potential of generative AI, and the human side of technology in the next era of CX.
The Chatbot Disappointment: Did Companies Learn?
At the start of 2024, Forrester warned that poor chatbot experiences were eroding customer trust. Have organisations gotten any better at deploying them?
“They’re still primitive,” said de Quintanilha. “Most chatbots today are helpful for a handful of tasks at best. Many misunderstand intent, and when they do respond, they often miss the mark.”
de Quintanilha points to a mismatch between expectation and delivery. Consumers want seamless, intelligent, language-based interactions, but are instead met with canned responses and clumsy logic trees.
The gap is a strategic misstep, rather than just a UX flaw. Generative AI (genAI) tools have enormous potential, but only when paired with thoughtful design, rigorous data, and clear use-case boundaries.
“Today’s generative bots are still too error-prone to fully lead customer interactions. But they’re useful in ‘whispering’ roles, suggesting solutions to agents who have the judgment to reject what doesn’t fit,” added de Quintanilha.
AI as Augmentation, Not Automation
While much of the industry continues to chase efficiency metrics, the real AI opportunity lies in augmenting, not replacing humans. de Quintanilha points to genAI’s behind-the-scenes role in helping service agents digest large volumes of information and provide more personalised, relevant support.
“We’re seeing agents get faster at answering questions, solving problems on the first contact, and leaving customers feeling heard and respected. AI takes the grunt work out of searching for answers. It frees people to be more strategic,” de Quintanilha said.
The same is true across internal operations. GenAI can summarise meetings, surface action items, and cross-reference knowledge from past projects, all in a fraction of the time it once took.
From Call Centres to Journey Orchestration
Where is AI actually delivering transformative results? “Journey mapping, journey analytics, and journey orchestration are going to be massively changed by generative AI,” said de Quintanilha.
Rather than treating journeys as rigid, step-based funnels, AI enables them to evolve in real time, matching the pace of customer behaviour.
Take CSG’s “Bill Explainer,” for example. Designed to simplify confusing telco bills, the tool uses genAI to provide personalised monthly bill breakdowns. The result is fewer service calls, faster payments, improved NPS, and a better experience for both customers and staff.
“It’s not just a digital tool. It’s also improved call quality, agent handle times, and payment speed,” she adds. “They’re now expanding it beyond telecom.”
In another case, a media company slashed video feedback analysis time from months to minutes and saw a 15-point NPS jump. And one CX vendor is training AI to mine transcripts for signals, like technical glitches or behavioural friction, that even agents may miss.
Data: More Than Just Integration
Despite these advances, there’s one persistent obstacle: data. “Connecting siloed data has long been a CX ambition. AI isn’t magic. You still need clean, reliable, interpretable data, and you need people who know how to use it,” said de Quintanilha.
Many organisations lack the AI literacy, internal governance, and platforms to make AI actionable. The solution isn’t more dashboards. It’s developing centres of excellence, investing in training, and building the right foundations, from responsible AI protocols to cross-functional collaboration.
Keeping CX Human in an AI Age
With AI embedded deeper into every touchpoint, is there a risk of customer experience losing its human touch?
“There’s always that risk. A lot of CX work is still based on personas that feel formulaic or disconnected from reality. AI doesn’t have to dehumanise, it can actually bring more context and richness to the table,” said de Quintanilha
In her view, the best outcomes will come from blending human judgment with machine intelligence.
“It’s the fusion of humans and AI that will define the next era of customer experience. Leaders who can orchestrate that blend, who know when to lean on tech and when to let people shine, will be the ones who create value customers actually feel,” she concludes.