Skip to content
Let's talk

Futurum Research

Insight · Market Coverage

AI Platforms8 min read

Zendesk CX Trends 2027: Ready to Manage an AI Workforce?

Zendesk’s Annual Survey Looks at How Companies Will Manage, Price, and Verify AI Agents in Customer Service

Vice President & Research Director, Enterprise Software & Digital Workflows

The short answer

Keith Kirkpatrick, VP, Research at Futurum, shares his insights on Zendesk CX Trends 2027 and what companies need to manage, price, and verify AI agents in customer service.

Futurum's Keith Kirkpatrick,

Cite this

Keith Kirkpatrick, The Futurum Group, "Zendesk CX Trends 2027: Ready to Manage an AI Workforce?," November 19, 2026. https://trial.futurumgroup.com/insights/zendesk-cx-trends-2027-ready-to-manage-an-ai-workforce/

Zendesk CX Trends 2027: Ready to Manage an AI Workforce?

Zendesk’s CX Trends 2027 report finds that 74% of business leaders surveyed expect AI agents to outnumber people in service interactions within two years. The same report shows that few companies have the management, pricing, and identity controls needed to run that many AI agents.

What is Covered in this Article

  • Key findings from Zendesk CX Trends 2027, including the 74% of leaders who expect AI agents to outnumber people in service interactions within two years.
  • Futurum data showing that enterprises are deploying AI agents in internal functions before customer-facing ones.
  • The gap between buyer demand for outcome-based AI pricing and what vendors offer today.
  • The lack of a way to verify AI agents that act on behalf of customers.
  • Why companies need to redesign service processes to manage AI agents well.

The News: On November 17, 2026, Zendesk released Zendesk CX Trends 2027, its annual customer experience report. Zendesk says companies are entering “the era of AI Management,” in which they manage AI as a workforce instead of deploying it as a tool. According to the report, 74% of business leaders surveyed expect AI agents to outnumber people in service interactions within two years, 81% already consider AI agents part of their workforce, and 87% say managing an AI workforce will be as important as managing human teams. Also, 90% say success depends on orchestration that connects agents, data, and systems, and 85% say AI vendors should price on verified outcomes instead of interactions processed. Zendesk ran two surveys in July and August 2026 across 10 countries. They covered more than 4,400 business respondents, from small businesses to enterprises, and more than 3,100 consumers.

Zendesk CX Trends 2027: Ready to Manage an AI Workforce?

Analyst Take: The 74% figure from Zendesk CX Trends 2027 will get most of the attention, but it tells buyers little. It is a forecast from a vendor-run survey of leaders who already use AI in service, and it counts interactions instead of measuring whether customers got their issues fixed. Other findings in the report are more useful because they show what companies still lack. Only 11% match AI capability to task complexity today. Also, 78% say coordinating several specialized agents on one customer issue is their biggest operational challenge, and 78% say they have no way to verify the identity of AI agents acting for customers.

Companies are adding AI agents faster than they are building the systems to manage them. Zendesk CX Trends 2027 shows gaps in orchestration, measurement, pricing, and trust, and I expect those gaps to slow adoption more than any limit in the AI models.

Enterprises Are Deploying AI Agents Internally First

Zendesk surveyed people who run service operations, so it makes sense that they expect AI agents to take over service work quickly. Enterprise buyers as a whole are putting agents somewhere else first. In Futurum’s 2H 2026 Enterprise Applications Decision Maker Survey (N=833, published August 2026), IT operations and observability (21.6%) and cybersecurity (16.4%) were the top first-ranked areas for agentic AI deployment. Sales, marketing, and service (12.8%) and customer engagement (10.4%) ranked lower. Enterprises are testing agents in internal functions, where mistakes are easier to catch and contain.

Zendesk’s forecast may still hold for digital-first brands and for the small and midsize businesses in its sample. These companies handle many simple, repetitive requests that AI agents can resolve today. The two-year timeline is less likely for complex issues that span several systems, and those issues cost the most to handle. Zendesk itself has forecast that AI agents will handle more interactions than humans in 2026. The 74% figure confirms a trend among early adopters; it does not show that the broader market has reached that point.

Zendesk CX Trends 2027 Shows Strong Demand for Outcome Pricing

In the report, 85% of leaders say AI vendors should price on verified outcomes. Zendesk introduced outcome-based pricing for AI agents in 2024 and expanded it with its Resolution Platform at Relate 2026, so the finding supports its own strategy. Buyers say different things when they have to choose a model. In Futurum’s 2H 2026 survey, among buyers who pay for generative AI as an add-on (n=645), 21.1% prefer outcome-based pricing and 36.6% prefer consumption-based pricing. Only 11.2% of those buyers say their preferred vendor offers outcome-based pricing today, and outcome-based pricing shows up in 8.6% of vendor pricing models overall.

Buyers want pricing tied to results, and few vendors offer it. That helps Zendesk and other vendors that already charge per resolution. Verification is the obstacle. Buyers and vendors have not agreed on what counts as a resolution, how to treat a resolution that leads to a repeat contact, or who checks the numbers. The report also finds that 86% of leaders want real-time visibility into cost per resolution. Buyers want to see the numbers first. Outcome pricing will spread only when both sides trust how resolutions are counted.

Companies Cannot Yet Verify AI Agents That Act for Customers

In Zendesk CX Trends 2027, 78% of business leaders say they have no solution for verifying the identity of AI agents acting for customers. I expect this to cause the most problems over the next 24 months. Meanwhile, 48% of consumers say they want to resolve service issues through a personal AI assistant without contacting the company directly, and 77% of leaders worry that AI systems will skip their brand if those systems cannot access the brand’s content.

Zendesk addresses part of this problem with its Open Service Directory, which helps AI assistants find a brand’s verified support channel. That helps the assistant find the right company. It does not tell the company whether an incoming agent has the customer’s permission to change an account, issue a refund, or access personal data. Consumers also give mixed answers: 71% would accept more friction for stronger fraud protection, but 75% are uncomfortable with behavioral analysis used for verification without their consent. Until standards for agent identity and permissions mature, I expect enterprises to limit what customer-side agents can do. Those limits will slow the growth in AI-handled interactions that the 74% forecast assumes.

Managing AI Agents Requires Process Redesign

Zendesk CX Trends 2027 is right that managing AI agents will become a core part of running service operations. In the report, 84% of leaders say AI workflow design is becoming a core skill for CX operations roles, and 64% expect to match AI capability to task complexity within 12 to 24 months, compared with 11% today. Companies will not close that gap by adding more agents to existing processes. As I wrote in Futurum’s research on why agentic AI’s real test is process redesign, companies get results when they rebuild workflows around the tasks agents handle well and the tasks people handle well.

I expect AI workforce management to become its own software category by the end of 2027, in the same way that workforce engagement management grew up around human agents. It will cover agent orchestration, quality monitoring, performance scoring, cost-per-resolution reporting, and governance. Zendesk can compete with its Resolution Platform and the agent technology it gained from Forethought. Salesforce, ServiceNow, Microsoft, Genesys, Sierra, and Decagon will compete too. Buyers should favor the vendors that make it easiest to measure and control AI agents.

The consumer data in Zendesk CX Trends 2027 also shows that 88% of consumers want a human available if needed. In the announcement, Shashi Upadhyay, Zendesk’s President of Product, Engineering and AI, says leaders need to decide where human judgment is needed. Using the 74% figure to justify staff cuts is likely to add AI agents without improving service. The better use of the finding is to redesign service operations around clear roles for AI agents and people.

See the complete press release on the findings of Zendesk’s annual customer experience report on the Zendesk website.

What to Watch

  • Whether Zendesk and its competitors publish audited autonomous resolution rates and cost-per-resolution figures, so buyers can test the forecast against actual results.
  • Whether buyers and vendors agree on what a “verified outcome” is, including how repeat contacts, escalations, and partial resolutions count in outcome-based contracts.
  • Adoption of standards for agent identity, permissions, and discovery, including whether other CX platforms and AI assistant makers support Zendesk’s Open Service Directory.
  • How Salesforce, ServiceNow, Microsoft, Genesys, Sierra, and Decagon respond as tools to supervise, orchestrate, and check the quality of AI agents become buying criteria.
  • Consumer and regulatory pushback on AI-based identity verification, since most consumers surveyed are uncomfortable with behavioral analysis used without their consent.

Sources

1. AI is reshaping service: 74% of business leaders surveyed ..., Zendesk



Declaration of generative AI and AI-assisted technologies in the writing process: This content has been generated with the support of artificial intelligence technologies. Due to the fast pace of content creation and the continuous evolution of data and information, The Futurum Group and its analysts strive to ensure the accuracy and factual integrity of the information presented. However, the opinions and interpretations expressed in this content reflect those of the individual author/analyst. The Futurum Group makes no guarantees regarding the completeness, accuracy, or reliability of any information contained herein. Readers are encouraged to verify facts independently and consult relevant sources for further clarification.

Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.

Read the full Futurum Group Disclosure.


Other Insights from Futurum:

Zendesk Launches Outcome AI Platform and Zendesk Unlimited Can Zendesk Employee Service AI Agents Pry IT Work From ServiceNow? Will AI Assistants Use Zendesk's Open Service Directory?

Published by Futurum.

About the author

Keith Kirkpatrick

Vice President & Research Director, Enterprise Software & Digital Workflows

Keith Kirkpatrick is VP & Research Director, Enterprise Software & Digital Workflows with The Futurum Group. His coverage area encompasses enterprise applications, platforms, and tools, and the technologies and tools that underpin them, including artificial intelligence, automation, and integration technologies. He has more than 25 years of experience in research, marketing, and consulting-based fields, and is frequently sought out by the vendor community and trade media for his insights and commentary.

Coverage on Zendesk
  • Analyst Insights51
  • Research Reports2
All Zendesk coverage
See the full picture in the platform

Vendor scoring, the benchmark data behind these findings, and the analyst team that produced them.

Open the platform (opens in a new tab)Become a client
The Weekly

Subscribe to The Weekly

The week's market signals, one email, no filler. Unsubscribe any time.