Expert Insights

Why the Smartest Contact Centres Learn From Every Conversation: Zoomtopia EMEA 2026 Highlights

Published On: 11th Nov 2026 08:00

Zoomtopia EMEA 2026 brought Zoom’s CX leaders, product teams and customers together for a live showcase of where its contact centre platform is heading next.

Across sessions on Zoom Contact Center, Zoom Virtual Agent, CX Insights, workforce management and customer journey orchestration, the event offered a useful snapshot of where AI-powered CX is right now: moving beyond isolated tools and towards systems that can learn from what customers actually do, say and need.

For contact centres, the next competitive advantage may not come from adding another AI feature. It may come from making better use of the thousands of conversations already happening across calls, chats, emails, virtual agents and back-office teams, then feeding those lessons back into the operation.

Zoomtopia showed how Zoom is trying to make that happen, alongside real customer examples of what the approach is already delivering.

Here are the biggest takeaways from the event, and what contact centre leaders should be paying attention to next.

Every Conversation Is Becoming Part of the System

Michelle Couture, CX and go-to-market product marketing lead at Zoom, started with a problem most CX leaders will recognise immediately: businesses have plenty of customer data, but far less continuity.

A support platform tracks a contact centre interaction. A CRM holds a case record. Another database contains a sales chat from three weeks ago. The issue is that critical customer signals often happen before those traditional workflows even begin.

According to research Zoom presented during the session, 9 out of 10 CX leaders said connecting conversations to outcomes was their biggest challenge, ahead of cost, headcount and tooling.

Two companies can use similar automation technology and get very different results if one is feeding it only contact centre transcripts, while the other is learning from sales calls, customer service, store interactions and back-office activity.

Zoom’s answer is a connected layer spanning those conversations, with Zoom Virtual Agent, customer profiles, CX Insights and workflow orchestration all drawing from and contributing to the same picture. As Couture noted: “Every resolved outcome feeds back into the system and makes it smarter for the next one.”

Adding another channel doesn’t simply give customers another place to ask for help. It gives the organisation another source of information about what customers want, where they struggle and whether the eventual solution actually worked.

And crucially, Zoom wants the system to improve as that information accumulates.

As Couture explained later in the session: “Every resolved outcome feeds back into the system and makes it smarter for the next one.”

Over time, those improvements compound as virtual agents learn from real customer interactions, supervisors catch issues earlier, and outcomes feed back into future interactions.

“It compounds, quarter over quarter, because the system itself was built to keep learning.”

Zoom Wants CX Intelligence to Lead Somewhere

Another standout theme from Couture’s session was simple: “Signals without action are just noise.”

It’s easy for contact centres to accumulate dashboards. It’s harder to make that information useful before someone spends an hour interpreting it.

Zoom introduced new CX Insights capabilities designed around that problem.

The Supervisor Alerts Hub can look across quality scores, workforce adherence data and conversation transcripts to surface patterns and recommend an action.
Scheduled Reports lets leaders request recurring analysis in natural language, while Executive Briefs turns results into presentation-ready outputs.
Customer Workflow Orchestration then extends the idea beyond reporting. If a customer agrees to return a product, for example, the process doesn’t necessarily need to die when the call ends and become somebody else’s ticket to remember. Zoom demonstrated a workflow remaining open until the warehouse confirms the return, at which point the next action can continue automatically.

Combined, these new features expressed a unifying theme: intelligence should shorten the distance between knowing something and doing something about it.

★★★★★

Case Studies: Employing AI in Action (that Customers Never Have To Think About)

Later in the event, Jody Rubin, Senior Product Marketing Manager for Zoom Contact Center, hosted Real Brands, Real Outcomes, bringing in organisations already using the technology in very different environments.

Higginbotham

Francine Thibault, Director of Employee Engagement at Higginbotham, described an operation that handles around 18,000 calls within a 45-to-60-day period during open enrolment.

Higginbotham’s approach was to take on operational foundations by rolling out AI Expert Assist and CX Insights. Call abandonment fell 64%, from 20.56% to 7.48%. Average handle time dropped 24.8%, wrap-up time fell 33%, and the organisation calculated approximately 1,693 labour hours and $42,325 in labour costs saved annually.

More importantly, Thibault said reports that previously took one or two hours to produce could now be answered in seconds.

“The biggest benefit isn’t the time saving, though,” she explained. “It’s that we’re spending more time acting on insights and improving the experience.”

Higginbotham also used AI to queue voicemail enquiries in the order they arrived. During peak periods, Thibault said the organisation had previously averaged between 50 and 100 return calls. That had fallen to around three at the end of each day.

Her conclusion could easily have been the tagline for the entire event, too, as she said: “The best AI is the AI customers never have to think about.”

Oxfordshire County Council

Another example comes from Oxfordshire County Council, where AI is helping to reduce the number of calls being passed around between departments.

During a recent CCH site visit, we heard how the council had replaced its traditional IVR with a voice AI assistant that can understand why residents are calling and direct them to the right team.

The impact was almost immediate. Calls reaching the switchboard fell by 55% on the first day, with that figure since improving to almost 60%. This has freed up advisors to spend less time transferring calls and more time dealing with enquiries that need their attention.

The council has also introduced Zoom Quality Management to automate parts of its quality assurance process, identify recurring customer issues and give team leaders more time to support their advisors.

What’s interesting about Oxfordshire’s approach is how much of the benefit comes from tackling fairly routine operational problems. Getting residents through to the right people first time might sound straightforward, but reducing unnecessary transfers can make a considerable difference to both the customer experience and the workload of the team.

We covered the council’s approach in more detail during our recent visit to Oxfordshire County Council

Robinson & Henry

Another CX practitioner, Jason Eversold, Director of Operations at Colorado-based law firm Robinson & Henry, had a very different environment to manage.

Its contact centre handles legal intake involving matters such as divorce and custody, where speed matters but accuracy, compliance and empathy matter just as much.

Eversold chose to use AI Intent Routing to understand what callers are actually asking for rather than forcing them through endless “press one, press two” menus.

AI Expert Assist was then implemented to guide agents through the relevant questions, compliance and next steps during the conversation, which Eversold described as being “like having a coach in their ear the whole time”.

The firm has reduced handle time from 12 minutes to seven minutes, while virtual-agent capabilities helped it extend assistance outside traditional office hours.

Eversold had also planned to recruit a workforce analyst in 2027, but CX Insights had effectively replaced that planned hire, because he can now ask operational questions directly and receive the information quickly enough to act on it.

The point here though isn’t that every contact centre should remove a future job from its recruitment plan – It’s that the value of AI is starting to show up in places beyond customer-facing automation, including forecasting, coaching, reporting and everyday management decisions.

★★★★★

Agent Architect: Where the Compounding Truly Starts

Arvind Rangrajan, Product Marketing Lead for Zoom Virtual Agent, brought the learning-loop idea together in the later Meet the AI Agent Behind Modern CX session.

The customer-facing demonstration was polished: a virtual agent identified a fictional customer, found his Shopify order, analysed a photograph of a damaged camera lens, arranged a replacement through FedEx and handed the conversation to a human agent with the context intact.

The most interesting part of that process all happened behind the scenes.

Zoom introduced Agent Architect: a tool that ingests batches of real support conversation transcripts to automatically build a virtual agent grounded in actual customer interactions rather than generic scripts.

Zoom’s Agent Architect was shown ingesting a batch of real support conversations and using them to identify common intents, required tools, knowledge sources and guidance for a new virtual agent. It then generated its own test scenarios.

The first simulation had returned a 97% resolution rate. Agent Architect identified why several previous tests had failed, proposed a fix to the email-collection process required for the Shopify lookup, ran the tests again and reached 100% in the demonstration before the agent was published.

Historically, building and maintaining virtual agents has been an exhausting, trial-and-error administrative chore. Agent Architect flips that entire paradigm on its head. By learning straight from real conversations and stress-testing itself in a simulated environment before a single customer ever interacts with it, the burden of maintenance plummets.

It transforms AI from a static project that requires constant manual babysitting into a self-improving, autonomous asset – meaning your virtual agents get smarter, sharper, and more reliable right out of the gate.

★★★★★

Final Thoughts

Zoomtopia EMEA 2026 made a compelling case for shifting how organizations view contact centres: moving away from static customer service silos and toward dynamic, learning ecosystems.

Every conversation can expose a knowledge gap.
Every resolution can improve the next interaction.
Every quality issue can become coaching data.
Every virtual-agent failure can become a test case.
And every operational pattern can potentially reach a manager quickly enough for them to do something useful with it.

The technology still has plenty to prove in messy, real-world environments, and a polished demo will always behave more elegantly than the average Monday morning contact queue. But Zoom also brought customers with numbers, lessons and working use cases, which made the vision far more tangible.

Perhaps the best advice of the webinar came when the customer panel was asked what they’d tell CX leaders still evaluating their first AI move.

“Don’t wait for the perfect use case.” Thibault said. “Start with a problem that’s creating friction for your customers or your employees and solve it… Focus on outcomes first, and technology second.”

Eversold was even more enthusiastic: “Honestly, just start. Your competitors are not hesitating, and neither should you. I think you’ll learn more in the first 30 days of using it than 6 months of evaluating it.”

By Zoomtopia EMEA 2027, the most interesting organisations probably won’t be the ones with the longest list of bullet-point AI features.

They’ll be the ones with a year’s worth of evidence showing what their systems have learned, what their teams changed because of it and how much better the customer experience became as a result.

For those interested in exploring these themes further, Zoom is hosting its next CX Tech Talk on 17 November.

Keep an eye on Zoom’s upcoming events for more details.