Marketing and service are becoming one continuous journey
Generated by Klaviyo AI
S&P Global's Sheryl Kingstone shares data showing agentic AI has risen to the top of enterprise priorities in just two years, with marketing and customer service converging into one continuous, always-on experience. She maps the gap between ambition and actual readiness, and what stays uniquely human as agents take on more of the day-to-day work.
- Ambition outpaces architecture: 64% of organizations have a digital transformation strategy, but only 28% have data they can actually act on in real time.
- The biggest blocker is structural: Cross-department collaboration ranks as the top barrier to automation, well above budget.
- Marketing and service are converging into one journey: The same agent use cases now top the priority list for both teams, built on a shared customer profile.
- Judgment, taste, and empathy stay human: Agents take over sensing, drafting, and resolving, while people decide what's worth saying and set the guardrails.
- Digitally driven organizations are pulling ahead: Companies with strong data foundations and enterprise-wide process ownership are twice as likely to see real revenue impact from AI.
Two years ago, “agentic AI” barely registered in S&P Global’s B2C data. Now, it’s at the top of the list. "We're moving from AI that assists to AI that acts," said Sheryl Kingstone, research director for customer experience and commerce at S&P Global in her talk at K:BOS. "It’s the fastest I’ve ever seen technology climb."
The numbers back that up. According to S&P Global’s Voice of the Enterprise: Customer Experience & Commerce, Organizational Dynamics and Budgets 2026 survey:
- 69% of merchants expect AI agents to be transformative to their business within the next 3 years.
- 73% now rate deploying agentic AI across sales, service, and commerce a top priority, the highest-ranked customer experience initiative.
- 49% said agentic AI is the single most game-changing technology for 2026, outranking customer data platforms and generative AI alone by more than 10 percentage points.
AI is dissolving the wall between marketing and service into one always-on agentic customer experience, according to Kingstone. But getting there will take more than ambition. It will require the data foundation and organizational architecture to show a 360-degree customer view.
The agentic shift, quantified
Kingstone put a number on what's at stake. Customer experience quality alone drives or erodes more than $400 billion in annual revenue every year, according to S&P Global’s Voice of the Connected User Landscape 2025 survey.
Some of that comes down to straightforward friction. Kingstone pointed to $67 billion in savings available through smarter digital channel deflection, $35 billion in cross-channel buying options in-store and online, and $22 billion in incremental purchases that better AI-assisted product discovery could capture, all from the Voice of the Connected User Landscape research.
Consumers, for their part, are ready. 67% of consumers plan to use an AI assistant in some way this holiday season, according to Klaviyo’s 2026 State of the Holiday Buyer research.
The tasks consumers trust agents with most are the ones that reward monitoring and instant analysis rather than judgment, like tracking an order, catching fraud, and managing a subscription. Kingstone called these the foundation of trust that more complex agentic interactions will need to build on.
The converged journey in action
For years, brands have demarcated their operations. Marketing brings customers in, then customer service takes over. Kingstone said that model is over: "Service is the new marketing."
According to the Voice of the Enterprise: Customer Experience & Commerce survey, 70% of organizations now view a unified digital experience suite, one that spans marketing, sales, and service, as “very strategic.” Creating a single source of truth for customer data scored 8.35 out of 10 in importance.
The same agent use cases now top the priority list for marketing and service teams alike. 91% of executives see personalized journeys as a valuable application of generative AI, a use case usually filed under marketing. 90% say the same about AI agents for customer service. A single suite spanning content, commerce, marketing, sales, and service is now the stated goal.
This is already showing up in practice. A marketing agent and customer service agent working off the same customer profile, instead of two separate tools that happen to sit near each other.
Kingstone's data from the 2025 Voice of the Enterprise survey on department-level agent plans matched up. 65% of marketing teams plan to use agents for automated customer support, and more than 52% of contact center teams say the same. 54% of marketing teams want agents for marketing optimization, and 45% of sales teams do too. Neither team was waiting for the other's permission.
Kingstone grouped what's forming underneath that convergence into 3 layers she called the agentic CX stack.
- Unified intelligence: Agents draw on an AI-ready data foundation for identity resolution, real-time signals, and 360-degree profiles.
- Agentic orchestration: Agents decide, act, and resolve issues without a person kicking off every step.
- Autonomous engagement: Agents personalize, sell, service, and recover across every channel in real time, in an always-on, self-correcting loop.
The practical effect, Kingstone said, is that the customer journey stops being a sequence of handoffs. Teams used to assemble insights after the fact, from batch reports. They planned campaigns on a quarterly cycle. Service picked up a customer cold, with no memory of the marketing that brought them there.
In the model Kingstone described, agents read signals continuously against a live profile, they created content at the moment of intent, and resolved a service issue with full visibility into the journey that led to it.
"The journey doesn't just get faster," she said. "It stops being a sequence of handoffs." That's also why the org charts, metrics, and team boundaries built around those handoffs are the first things to break.
The gap between ambition and readiness
According to S&P Global’s Voice of the Enterprise: Customer Experience & Commerce, Digital Maturity 2025, 64% of companies have a formal digital transformation strategy. Far fewer, only 28%, have actionable, real-time, 360-degree customer data. Just 23% manage customer experience processes across the whole enterprise rather than department by department. And only 18% describe their operations as primarily automated.
"Ambition is close to universal," Kingstone said. "Architecture is not." Agentic AI, she added, amplifies whatever foundation it's built on, which is why organizations that aren’t set up with a solid data infrastructure produce more fragmented, and more frustrating, experiences.
The blocker to fully automating operations was not budget constraints, which ranked 11th on Kingstone's list of friction points, at just 19%. The top barrier, at 35%, was cross-department collaboration and operating model alignment. The number two barrier was rising customer expectations outpacing AI capability (34%), followed by privacy and AI transparency requirements (33%), data security and model governance (33%), and the total cost of running a sprawling CX stack (32%).
"The barrier to transformation is structural," Kingstone said. "We have every excuse not to make this work. We have to get over these constraints. You can't buy your way to a continuous journey. You have to build your way there."
In S&P Global’s Voice of the Enterprise survey, organizations she calls digitally driven, the ones with both a real-time data foundation and enterprise-wide process ownership, are more than twice as likely to report a strong positive revenue impact from generative AI. More than 35% of them are actively exploring agentic AI capabilities, compared with just 13% of digitally delayed peers.
“We need to take a leap of faith and follow digitally driven organizations’ lead,” said Kingstone. “Digitally driven leaders are exploring agentic AI capabilities, they’re responding and getting revenue impact from GenAI, and they have enterprise processes. Digitally delayed companies are having problems with their foundation, not focused on automation, and having a hard time delivering value.”
The indispensable human
None of this, Kingstone was careful to say, replaces the value people bring. It relocates it. From execution to governance, there’s a new role for humans to manage the “digital worker.”
According to the Voice of the Enterprise survey, 64% of leaders now rate AI governance and autonomy controls as a very strategic need, and 75% say building trusted customer relationships requires new investment in managing preferences, privacy, and security. Kingstone framed that as a shift from doing the work to setting the guardrails agents work inside of.
She drew a specific line between what agents are ready to own and what stays with people:
Agents are on track to own:
- Sensing, watching real-time profile and identity signals to catch anomalies and stop fraud
- Planning and drafting, turning campaign briefs into ad content, dynamic text, and images
- Sending and testing, running campaigns across email, mobile, and messaging, then optimizing from the results
- Resolving, working service issues through conversational agents, intelligent routing, and call summaries
- Recommending, surfacing products and next-best actions at the moment of interaction
What stays human:
- Judgment, making calls that carry commercial, legal, or reputational risk
- Taste and brand, knowing what is worth saying and what to reject
- Empathy, knowing when a customer needs to be heard, not just resolved
- Governance, setting autonomy boundaries, approval paths, and audit standards
- Framing, deciding what agents should optimize for in the first place
"The line moves," Kingstone said, "but it does not disappear."
The skills fading fastest in her data are the hands-on ones, like manual channel execution, routine ticket handling, static segmentation, and production work done at volume. In their place, Kingstone sees growing demand for agent supervision, context stewardship (shaping the data agents draw from), and governance and trust design.
Org charts are also changing. Execution-heavy layers are flattening while senior, strategic roles hold steady, and the metrics teams get judged on are shifting from raw output to agent ROI, quality, containment, and trust.
What CX leaders should prioritize for 2027
Kingstone closed with a clear path for CX teams in the era of agentic AI.
- Treat the data foundation as job one. Agentic AI amplifies whatever it's built on, and only 28% of organizations have an actionable, 360-degree view of the customer to build on.
- Fix the operating model before adding agents. The barriers cited most, cross-department collaboration and legacy application complexity, are structural, not financial.
- Design for one continuous journey. 70% of organizations call a unified experience suite “very strategic.” Stop optimizing marketing and service separately, and instrument the handoffs that agents will run end to end.
- Architect for governance from day one. Build in explainability and audit trails before scaling an agent, not after, given that 64% already call governance and autonomy controls a strategic need and 75% say trust requires new investment.
"Everyone thinks AI agents are revolutionary, but I think it’s evolutionary,” said Kingstone. “The shift to an agentic customer experience is the single biggest transformation our industry has ever faced. And it will be led by the people who are willing to move from ambition to architecture."