Talkdesk report says AI adoption outpaces business gains

Talkdesk has published a report on the gap between AI adoption in customer service and measurable business results. The study found that 98% of organisations have deployed AI across the customer journey.

Conducted by NewtonX for Talkdesk, the survey drew on responses from 252 senior decision-makers responsible for customer experience, IT, operations or AI strategy at mid-market and enterprise organisations across North America, EMEA, Latin America and Asia-Pacific.

The findings suggest that widespread use of AI tools has not translated into joined-up execution. Only 15% of organisations said they combine agentic AI with cross-departmental orchestration to resolve customer needs from start to finish, and just 5% said they can quantify AI's impact on business outcomes.

According to the report, 85% of organisations lack the orchestration needed to connect AI agents, human teams, data and workflows across enterprise systems. As a result, businesses pay for separate AI tools while still bearing the cost of unresolved customer requests and manual follow-up.

Execution gap

The research points to a series of operational weaknesses behind the gap. While 64% of organisations use specialised AI agents, only 35% said they retain customer context when moving from one system to another, making end-to-end resolution harder to achieve.

Technical barriers also remain common. Disconnected systems were cited by 45% of respondents and legacy infrastructure by 44%, while nearly 80% of organisations are limited to 10 or fewer AI automations.

When automated processes break down, the burden falls back on staff. According to the findings, human agents lose an average of 28% of their time switching between systems, re-entering data and searching for customer context.

Tiago Paiva, Chief Executive Officer and Founder of Talkdesk, said the numbers show a mismatch between AI activity and the organisational structures needed to produce clear returns.

"The findings expose a widening divide between AI activity and the operating capabilities required to deliver business impact. This disconnect creates a false sense of progress. Widespread AI deployment masks the reality that so few organisations can quantify AI's impact on business outcomes. As organisations shift from deploying AI tools to managing AI as part of their workforce, orchestration has become the defining capability separating experimentation from measurable business impact," said Paiva.

Hybrid workforce

The report also suggests some organisations are beginning to view AI less as software and more as part of the workforce. Nearly one in five respondents said they see AI agents more as labour than technology, while 99% said a hybrid workforce of AI and people creates value.

Even so, management models and governance appear to be lagging. Organisations with the highest level of customer experience automation maturity were more than 10 times as likely to run AI and people as a unified operation.

Trust remains a barrier to wider delegation. More than half of respondents, 52%, cited trust in AI decisions as a primary concern, and 94% said they operate without AI-assisted knowledge management.

That absence matters because knowledge management systems often provide the information foundation automated systems need to answer questions accurately and handle more complex requests without human intervention.

Measured gains

The report groups organisations along a maturity curve based on agentic AI use and cross-departmental orchestration. On that measure, companies with stronger integration were four times more likely to report major gains in customer satisfaction or Net Promoter Score.

Those with higher maturity were also nearly twice as likely to automate revenue-related use cases such as churn prediction and personalised recommendations. Among the leading organisations in the survey, 38% said they autonomously resolve more than 40% of customer issues, while none in the lowest maturity tier reached that level.

Zeus Kerravala, Principal Analyst at ZK Research, said the results show that adoption alone is not enough.

"Talkdesk's benchmark data is a wake-up call: deploying AI is easy, but operationalising it is where most enterprises stall. Moving from AI experimentation to real execution requires an operating model where AI, people, data, and workflows operate as a unified workforce. That's where true business value will be created," said Kerravala.

For Talkdesk, the findings support a broader industry argument that the next stage of AI in customer service will depend less on adding more standalone tools and more on linking systems, workflows and people in a coherent operating model.

Munil Shah, Chief Product, Technology, and Customer Officer at Talkdesk, said: "AI agents are becoming part of the customer experience workforce. But organisations need more than another AI tool - they need a way to manage how AI and people work together. We built Talkdesk Customer Experience Automation on the belief that AI creates value when it can coordinate work across systems, departments, and people. The research findings show why that orchestration matters: the companies seeing the greatest business impact aren't just adding more AI - they're orchestrating work across a hybrid workforce that delivers on the outcomes customers and businesses expect."