At Zendesk’s Relate 2026 conference, the message was clear: Firms must shift from deploying isolated AI features to operationalizing AI across the entire service lifecycle, or risk failing to realize measurable business outcomes. From platform breadth and integration to knowledge readiness and new service roles, the event highlighted that success with AI depends on much more than adding automation.

To understand Zendesk’s current strategy, it’s worth looking back four years. Since being taken private in 2022, Zendesk has completed over 15 acquisitions for over $500 million. These acquisitions have shaped its product into a comprehensive AI-first customer service suite.

Zendesk first rounded out its customer service platform by purchasing Klaus for QA and Tymeshift for workforce management, completing the suite with Local Measure for CCaaS. Zendesk strengthened its platform with Ultimate for AI agents, HyperArc for analytics, and Forethought for conversational AI. Expanding into the EX space, Zendesk also acquired Unleash, a search and knowledge retrieval solution.

Zendesk’s Resolution Platform Continually Optimizes Outcomes

Kate Leggett, VP, Principal Analyst

At the conference, Zendesk evolved its vision for its Resolution Platform, announced in 2025, to one that delivers customer outcomes. The platform now integrates AI agents, copilots, knowledge, workflows, and governance into a single system designed to resolve customer interactions and improve via a built-in learning loop. Highlights included: AI agents capable of handling multistep workflows across channels (messaging, email, voice) and across the front, middle, and back office; domain-specific AI agents, including employee service agents for IT and HR and agent builder tooling; role-specific copilots; and connectors and a context graph. Another defining element was Zendesk’s continued push toward outcome-based pricing, where customers pay for verified resolutions, which aligns commercial models with business outcomes.

The breadth of recent acquisitions is impressive, but it also results in many overlapping capabilities. Zendesk has yet to prove the true autonomy of its resolution learning loop. Keep an eye on how Zendesk brings these products together into a full suite rather than a “collection of tools” — differentiation on how effectively firms integrate capabilities into a unified platform rather than introducing a collection of innovations.

For Customer Service, It’s All About Breadth

Max Ball, Principal Analyst

The company delivered a strong, but not especially differentiated, set of AI capabilities across its ticketing system, CCaaS, self-service, and agent tooling. The real story isn’t the features themselves, which increasingly look similar across vendors, but Zendesk’s potential to bring them together on a single platform. Recent acquisitions of Local Measure for CCaaS and Forethought for customer self-service expand the company’s value proposition for customer service teams, improving its competitive footing.

The proof, however, will be in the integration; Zendesk must move beyond surface-level integration and truly embed AI across all its customer service capabilities. This provides the company with a path to stand out. For customer service teams, the company’s biggest advantage is the ability to unify rich service data — from tickets to interaction transcripts — into more personalized, context-aware experiences. With Zendesk’s reach beyond customer service into other areas of the business, the value of the customer data can be amplified across the entire customer journey.

Quantifying The Knowledge Gap Was The Easy Part

Julie Mohr, Principal Analyst

Tom Eggemeier, Zendesk CEO, introduced a metric at the Relate 2026 keynote that converts knowledge readiness from a qualitative argument with executive sponsors into a numeric one. Twenty-seven percent of knowledge artifacts were ready immediately; another 28% were “unlockable resolutions” blocked by a specific knowledge gap. Closing those gaps, however, requires more than better documents. Documentation written for a human reader who interprets ambiguity is not sufficient input for an agent that executes on what it reads. The operational discipline that turns an unlockable resolution into a live automation is converting policy from narrative form to encoded form, where parameters, conditions, fallbacks, and edge-case logic are specified rather than assumed.

Doing this work demands a role that few service organizations have yet staffed. Christina Diaz at Supercell named the shift in plain terms during the agentic CX transformation panel: Human service agents are becoming “service engineers designing and curating the work that the AI agents will then execute.” Supercell already restructured around the new operating model, with dedicated product management for AI customer experience, a renamed player journey team taking on the former process design function, a separate trust and safety function tracking regulation, and explicit upskilling of BPO partners toward AI service architect roles. This is a trend that Forrester is tracking.

Elymae Cedeño described the same pivot at Bumble, separating interactions suited to fast automation from those that require high-touch human contact. Forrester clients planning the next 12 months of their AI rollout need to staff for the new types of work required now, before automation crosses the inflection point. The window between recognizing the shift and staffing for it is where the best tier-one agents either become knowledge engineers or leave for organizations that already have made the role real. Knowledge engineering, not technology, is now the primary constraint on achieving scalable AI success.

Let’s Connect

Have questions? That’s fantastic. Let’s connect and continue the conversation! Please reach out through social media or request a guidance session. Follow our blogs and research at Forrester.com.