Customer experience has reached an inflection point. Generative and agentic AI are changing how people discover products, request services, make decisions, and interact with organizations. But the defining challenge is no longer whether enterprises can introduce AI into their business. It is whether they can redesign experiences, workflows, and operating models so that AI delivers outcomes people value.
This distinction shaped a recent discussion at ServiceNow Knowledge 2026, where leaders from Capgemini, ServiceNow, Sage, and DHL explored human-led, AI-powered transformation in practice. The conversation, moderated by Tim Arkin, DCX Global Head of ServiceNow at Capgemini, brought the findings of the Capgemini Research Institute AI-Powered-CX_April 2026 to life through client, technology, and transformation perspectives.
The message for business leaders is clear: success will not be determined by the number of AI agents an organization deploys. It will depend on how effectively AI is connected to the people, data, workflows, and governance required to complete work and improve experiences at enterprise scale.
The ambition is high, but the experience remains fragmented
The research, based on a survey of 9,500 consumers and 1,200 senior leaders and front-line staff across ten sectors, shows the pace of change. 68% of organizations believe AI agents could outperform traditional CX channels, while 20% have already enabled agentic AI in some or most CX activities. Additionally, 58%of consumers believe agents could save them time by automating routine monthly purchases and payments.
Yet, adoption alone does not automatically improve experience. An organization can automate an interaction while leaving the underlying journey fragmented. It can make one channel faster while simultaneously losing customer context at the next touchpoint. It can also introduce more agents and interfaces while making the overall experience more difficult to navigate
This execution gap is significant. 84% of executives view CX as a core driver of growth, but only 36% report improved sales after acting on CX feedback. Just 23% have a unified CX strategy across channels, 40% identify fragmented journeys as their greatest CX barrier, and 28% transfer context and conversation history seamlessly between online and offline channels.
This is precisely where partnership matters. ServiceNow provides the platform foundation to connect interactions, data, workflows, and outcomes. Capgemini brings the industry, process, experience-design, integration, and transformation expertise needed to turn that foundation into measurable business impact. Together, our shared ambition is to help the enterprise run as one, with AI integrated into workflows rather than added as another isolated channel.
Begin with the experience, not the AI agent
For Kerry Race, Global People Services Director at Sage, transformation starts by understanding employee journeys. Sage has introduced Now Assist into its people function, alongside a virtual agent intended to help colleagues access answers more efficiently. However, efficiency is only one part of the design.
For Kerry Race, Global People Services Director at Sage, transformation starts by understanding employee journeys. Sage has introduced Now Assist into its people function, alongside a virtual agent intended to help colleagues access answers more efficiently. However, efficiency is only one part of the design.
But the question is not how much of the journey can be automated, but where automation genuinely improves the experience. Routine enquiries can be solved digitally, while complex, sensitive, or emotionally significant solutions require human judgment, guidance, and trust.
Research reinforces this hybrid model. More than 66% of consumers rank front-line employees among their three most important interaction channels. At the same time, 76% of front-line staff say AI helps them spend less time on mundane activities and 70% report improved response quality.
The findings indicate that the most effective experiences require intentional automation. AI can streamline routine interactions and surface relevant information, while people remain focused on work requiring empathy, creativity, accountability, or nuanced decisions.
As Kerry Race advises, organizations should spend time with the people who will use the experience, understand their frustrations, and test whether the proposed solution meets an actual need. Leaders should not assume they already know the answer.
Put process and accountability before technology
The operational perspective from Frank Schüler, Global Managing Director at DHL, reinforces why platform-enabled transformation must begin with process.
AI is accelerating a digitalization journey that already includes deterministic process automation and robotic process automation. It adds valuable non-deterministic capabilities, but it does not eliminate the need for disciplined process design.
Frank Schüler describes the model simply: people lead, AI thinks, and robots act. Within this hybrid workforce, the responsibilities of people, AI agents, and conventional automation must be clearly delineated.
From there, organizations can make informed decisions about where AI belongs within a process. Some activities remain highly structured and rule-based, while AI is better suited to tasks involving unstructured data, information synthesis, or decision support. The goal should be an end-to-end journey where people, AI, and automation work together to deliver a desired outcome rather than forcing AI into every workflow.
At DHL, a holistic workforce-management approach built on ServiceNow Customer Service Management orchestrates services across human resources, RPA bots, and emerging agentic AI solutions. That model rests on three foundations: customer-focused processes, clear and reliable data, and well-defined interfaces between tools and resources. Governance, roles, responsibilities, and accountability complete the foundation.
Connect interaction, work, and outcomes
Ian Henderson, Senior Director, CX Ecosystem at ServiceNow, identifies three platform requirements for moving from disconnected insights to business outcomes.
First, connect the data. Customer-service organizations have historically separated interaction systems, workflow platforms, and systems of record. Unifying interaction data, work data, and outcomes creates a stronger foundation for AI.
Second, move from assistance to fulfilment. Generative AI can summarize and recommend. Agentic AI can take responsibility for completing work against a defined outcome. This changes the conversation from “Where can AI help?” to “What outcome does this person need, and how should human and AI resources deliver it?”
Third, make AI observable. Completing a task is not enough if the outcome is inconsistent, costly, or difficult to govern. Organizations need visibility into performance, failures, costs, and where benefits are accruing. Observability and an AI control tower provide the mechanisms to connect experimentation with operational and financial value.
This is the overarching Capgemini and ServiceNow perspective: govern with purpose, run as one through a unified platform, and continuously redefine value through human-AI collaboration. Capgemini helps redesign the organization and its processes, while ServiceNow acts as the enterprise nucleus connecting AI agents, shared data, and workflows.
Organizations should begin with controlled experimentation, route a limited share of interactions through new agent-driven processes, compare outcomes, and scale only when evidence supports it. The future of CX will not be defined by AI adoption alone. It will be defined by simpler journeys, connected work, accountable automation, and experiences that remain unmistakably human.
Read the full report here: Reimagining customer experience: Human-led, AI-powered CX transformation












