Customer Journey Orchestration Platforms

Forrester's Q2 2026 Landscape for this market identifies agentic AI as the top disruptor, and describes its effect with a word worth stopping on: compression.

Agentic AI compresses journeys.

For a category whose entire product is the journey, that is an unusual thing to have identified as the dominant force acting on it. Orchestration assumes a path with stages, moments of hesitation, and opportunities to intervene. A compressed journey has fewer of all three.

The same Landscape describes organisations moving from managing customer journeys to enabling AI to act on them autonomously in real time, while raising the bar on traceability, regulatory compliance, and outcome-driven decisioning.

That combination, shorter journeys and stricter accountability for the decisions made inside them, is the shape of this market now.

What orchestration does

Forrester's framing of journey orchestration is precise: using real-time, individual-level data to analyse behaviour and adjust the journey in the moment.

Three constraints sit in that sentence and each one is where implementations fail.

Real-time means the system notices while the customer is still acting. A payment fails, a cart is abandoned, a support article is read three times, a login attempt is repeated. If the system responds tomorrow, orchestration has become reporting.

Individual-level means the decision is about this person rather than about a segment they belong to. Segment-level personalisation was solved a decade ago and is not what this category sells.

And adjust the journey means acting, not recommending. The output is a routed call, a suppressed campaign, an offer, an intervention, a change to what happens next.

Forrester describes it as the nerve centre of a journey-centric technology stack, connecting journeys and supporting decision-making grounded in a customer's actual path and perceptions as they pursue a goal.

It sits alongside but distinct from journey management, which coordinates the organisation's improvement work across quarters. Orchestration operates in milliseconds against individuals. The two markets have almost no vendor overlap.

Inside The Forrester Wave: Customer Journey Orchestration Platforms, Q2 2024

The evaluation scored nine providers against thirty criteria: Alterian, CSG, Engage Hub, Genesys, inQuba, Medallia, Qualtrics, Quantum Metric, and Roojoom.

CSG placed as a Leader with the highest possible score in sixteen of the thirty criteria, including customer journey analytics, cross-journey orchestration, and real-time decisioning.

Quantum Metric appeared for the first time, and Forrester's characterisation of it is instructive. It described a newcomer used across retail, telecommunications, healthcare, B2B technology, and government to drive continuous product design from journey signals, democratising access to journey analytics including session replay. The vendor's own positioning is candid that it complements rather than replaces a conventional orchestration platform.

Look at where those nine come from and the category's shape becomes clear. Medallia and Qualtrics arrived from experience management. Genesys from the contact centre. CSG from billing and customer engagement in telecommunications. Quantum Metric from digital experience analytics. Alterian and Engage Hub from marketing communications. inQuba and Roojoom as journey specialists.

Eight or nine different starting points, which tells you orchestration is a capability that several adjacent markets concluded they needed rather than a market that formed on its own.

Forrester added new criteria for that edition covering predictive journey AI and generative AI, noting that predictive AI and machine learning had been core criteria since the inaugural 2018 evaluation but that generative capability was now being incorporated by every vendor, with some building standalone products around it.

What customers said was hard

Forrester recorded two challenges from reference customers, and neither is about orchestration logic.

The first is integrating data sources into a real-time unified view of the actual journey, so that issues across channels surface and opportunities can be acted on.

The second is hyperpersonalising increasingly conversational journeys for customers whose behaviour keeps changing.

The first challenge is the one that decides whether a deployment works. Orchestration requires the system to know, right now, what this specific person has done across every channel. Most organisations hold that data in a marketing platform, a service platform, a product analytics tool, and a system of record, each with its own identity resolution and its own latency.

Each of those views can be internally correct and mutually contradictory. The customer who complained on Tuesday, browsed on Wednesday, and called on Thursday exists as three separate people until something reconciles them, and reconciliation after the fact is not real time.

That is why orchestration platforms increasingly depend on a customer data platform underneath rather than solving identity themselves. It is also why the honest sequencing question for a buyer is whether the data foundation exists, because an orchestration engine on fragmented data orchestrates confidently against a partial picture.

What compression actually means

Return to the 2026 finding, because it changes the category's premise rather than its features.

A traditional customer journey has length because the customer has to do work. Search for options, compare them, read reviews, visit a site, hesitate, return, abandon a cart, receive a reminder, come back, convert. Every one of those steps is a moment where an orchestration platform can observe and intervene.

An agent acting for the customer removes most of them. It gathers options, evaluates against stated criteria, and transacts. There is no hesitation to detect, no cart to abandon, no reconsideration window to influence. The touchpoints that orchestration was built to work with collapse into a smaller number of machine-mediated interactions.

That does not eliminate the category. It changes what it orchestrates. The remaining moments become more consequential precisely because there are fewer of them, and the interaction shifts from persuading a person to satisfying an agent's evaluation criteria.

It also revalues the parts of the journey that agents do not compress. Complaints, service failures, complex decisions with emotional weight, and anything requiring negotiation or judgement remain human and remain long. Orchestration's centre of gravity moves toward those, which is closer to service recovery than to conversion optimisation.

Any evaluation running now should ask vendors directly what their platform does when the counterparty is software rather than a person, because the criteria in the current research were not written for it.

Traceability becomes a requirement

The other half of Forrester's 2026 finding deserves as much attention as the compression half.

Raising the bar for traceability and regulatory compliance follows directly from autonomous decisioning. When a human designs a journey flow, the logic is inspectable: someone drew it, someone approved it, and the path a customer took can be reconstructed from the diagram.

When a model decides in the moment what a specific individual should receive, that reconstruction is harder, and the requirement to produce it is rising rather than falling.

In several sectors this is not optional. Financial services decisions about offers, pricing, and eligibility carry regulatory expectations about explanation and fairness. In Europe, automated decision-making that produces significant effects on individuals sits under specific GDPR provisions, and the EU AI Act adds transparency and documentation obligations for certain systems.

The practical requirement for a platform is being able to answer, months later, why this customer received this treatment at that moment, with the inputs and the rule or model version recorded. That is an architectural property, not a reporting feature, and vendors differ substantially in whether it was designed in or added later.

EVAM's recognition in the Q2 2026 Landscape reflects this pairing, cited for agentic journey orchestration alongside outcome accountability, with strength in financial services, retail, and telecommunications across EMEA. Those are the sectors where autonomous decisioning and audit requirements collide hardest.

Two kinds of agent, pulling opposite ways

There is a distinction worth keeping straight, because vendors blur it.

Agents acting for the brand are what most orchestration vendors mean by agentic AI. The platform gains autonomy, deciding and acting without a human designing each path. That is an extension of what these systems already do, with more decision authority and less predetermined logic.

Agents acting for the customer are the disruption Forrester named. They change who the platform is interacting with, and the entire behavioural model underneath orchestration assumes a human: attention, hesitation, susceptibility to framing, response to urgency.

A platform optimised to detect a human hesitating will read an agent's systematic evaluation as something else entirely. Behavioural signals that indicate a wavering buyer indicate nothing when the counterparty is software methodically working through options.

Most vendor roadmaps address the first and are quiet on the second. That is understandable, since the first sells and the second is a threat to the premise. But the second is what Forrester identified as the top market disruptor, and a buyer signing a multi-year agreement should ask about it explicitly.

Where this leaves the category

Journey orchestration remains the mechanism by which an organisation acts on what it knows about a customer in the moment it matters, and nothing has replaced that function.

What has changed is that the moments are becoming fewer and more consequential, the decisions are becoming more autonomous, and the accountability for those decisions is becoming stricter. Those three trends do not point in the same direction, which is the interesting tension in this market.

Fewer moments argue for higher stakes per interaction and therefore more sophisticated decisioning. Stricter accountability argues for decisioning that can be explained, which usually means simpler. Autonomous action sits between them and increases the cost of getting either wrong.

The vendors that resolve that tension well will be the ones that can act autonomously and account for it afterwards. On current evidence, most can do one or the other.

Forrester does not endorse any vendor named here, and tier placement should not be read as a recommendation to buy.