Customer experience (CX) has long been a durable source of competitive advantage: organizations that lead in CX tend to outperform peers on growth and shareholder returns by deeply understanding customers, redesigning end‑to‑end journeys, empowering frontline teams, and rigorously managing performance.
Today those capabilities are becoming table stakes. Generative and agentic artificial intelligence (AI) can be a tool to build persistent advantage, but many organizations have seen only limited results because they layer AI onto fragmented, static journeys—optimizing steps rather than the live decisions that shape what customers actually experience.
From journeys to decisions in motion
Traditional customer journeys are static maps of intended paths. By contrast, "decisions in motion" are the real‑time choices that determine what an individual customer sees and receives: which offer appears, which policy applies, what help is triggered, and when a human should step in. Agentic AI enables goal‑driven agents to interpret context continuously, resolve ambiguity, and decide when and how to act across channels and operational systems. Human judgment remains essential but shifts upstream to set objectives, guardrails and escalation points while agents handle moment‑to‑moment execution at scale.
This is a movement from static optimization to dynamic orchestration.
Why CX is emerging as a proving ground for agentic AI
McKinsey data show that 41 percent of AI deployments in customer‑facing functions (including personalization, operations optimization, and contact‑center automation) have fully scaled—about 3.5 times the likelihood of scaling versus other business domains. In practice, CX agents often reach production faster, deliver measurable impact, and sustain adoption.
A key reason is structural: over the past decade CX‑focused organizations have invested heavily in digital infrastructure, journey redesign and performance management, creating the necessary foundations for agentic systems—observable workflows, explicit decision rights, structured data and human‑in‑the‑loop models. CX therefore serves as both a practical launchpad for organizations early in their AI journeys and a high‑impact arena for mature enterprises to move from isolated use cases to coordinated, workflow‑level autonomy.
Three horizons of agentic CX
Progression typically follows increasing decision authority and coordination scope:
-
Horizon 1 — Workflow rewiring: Agents autonomously execute a single, well‑defined workflow end to end under strict guardrails and escalation rules. Most current agentic CX production deployments sit here because value is clearest and the path to scale is shortest. Success requires clean process design, structured inputs, explicit business rules, defined escalation paths, and human‑in‑the‑loop controls from day one.
-
Horizon 2 — Domain orchestration: Agents coordinate multiple workflows inside a CX domain, optimizing decisions across workflows so handoffs preserve context and the customer experience feels continuous rather than fragmented.
-
Horizon 3 — Ecosystem experience engine: Agents orchestrate decisions across functions, channels and partners against shared objectives, optimizing the full end‑to‑end experience rather than improving isolated silos.
Workflow rewiring is necessary but not sufficient: automating a single workflow without clarifying the underlying decision layer—what agents may decide, what context they must use, when to escalate, and how decisions link to adjacent workflows—risks hard‑coding fragmentation into a faster system. Horizon 1’s value comes from demonstrating decisions can be made consistently, transparently and safely within clear boundaries.
Examples and measurable impact
Agentic solutions add value quickly when anchored in high‑volume, repeatable moments. McKinsey mapped 17 core CX workflows and identified six priority workflows that are strong near‑term candidates because they contain high‑stakes, repeatable decisions that affect both customer outcomes and cost.
One concrete example: a large European telecommunications provider with roughly €40 billion in revenue built predictive models that link signals across domains and channels to decide not only which offer to present but when, where and with what supporting actions (for instance, a proactive service fix or eligibility check). The integrated system delivered a €40 million margin improvement through higher engagement and conversion—an illustration of how outcomes improve when decisioning connects across the lifecycle instead of optimizing a single channel.
Consumer behavior is already shifting: a recent McKinsey survey found 44 percent of European consumers who tried AI search now use it as their primary search method, and about half of consumers use AI when searching for products online—signals that agentic commerce and coordinated decisioning have momentum.
Moving to ecosystem‑level coordination and the risks involved
True transformation occurs when workflows share context, objectives and memory so decisions in motion can be optimized across the full customer lifecycle. Horizon 3 requires more than scaled use cases: it requires a shared decision layer with persistent customer identity and unified context, clear enterprise objectives balancing customer value, cost, risk and capacity, and agent abilities to act across systems under strong governance, standards and auditability.
Because agents act across systems, data and workflows, small errors—pulling the wrong record, misapplying a rule, triggering an incorrect escalation—can compound into privacy breaches or broken customer promises. CX is therefore higher‑stakes than many other functions: speed and personalization must be matched by control. Agentic CX systems should be designed with:
- clear human and agent roles and decision boundaries,
- defined escalation paths,
- continuous monitoring and feedback,
- strong identity, access and audit controls to keep actions transparent and reversible.
Organizations that embed these safeguards early can scale autonomy while protecting customer trust. McKinsey’s Rewired framework explores these elements in greater depth.
Conclusion for leaders
Agentic AI will redefine how customer experience is delivered: as decision‑orchestration rather than static journey design. Companies that embed agentic capabilities into their operating models will accelerate learning cycles, respond with greater precision and redeploy human talent toward higher‑order judgment and relationship building.
The key choice for leaders is not whether agentic AI will reshape CX but how deliberately they will redesign around it. Treating agentic AI as just another technology layer yields incremental gains; redesigning CX around governed "decisions in motion" can deliver more coherent, responsive, personal and trusted customer experiences.
Authors and editorial
Authors: Alex Rodriguez (senior partner, McKinsey Miami office), Nicolas Maechler (senior partner, McKinsey Paris office), Ema Karavdic (associate partner, McKinsey New York office) and Khalid Quidwai (associate partner, McKinsey New York office). Edited by Larry Kanter (senior editor, McKinsey New York).
