Zoom Resolves 97% of Customer Queries with Its Own Virtual Agent

The enterprise communications giant shares its experience in implementing the Zoom Virtual Agent across its contact center operations

4
Sponsored Post
Zoom Resolves 97% of Customer Queries with Its Own Virtual Agent
Contact CenterConversational AIInsights

Published: February 20, 2025

Charlie Mitchell

Sometimes, a customer experience vendor will wax lyrical about the benefits of a solution but not implement it across its own operations.

There’s no better example of this than with conversational AI, as many providers haven’t put their virtual agents front and center on their websites.

Even when they do, they’re often restricted, decision-tree bots, not the natural language agents they sell to their customers. Cue alarm bells.

Zoom is different. Go to its website, and its virtual agent is there. It responds in natural language, cites its source content, and provides both rule- and generative-based conversational experiences (more on this later!).

Importantly, it also escalates to a live agent upon request, who receives a summary of the conversation so far and can pick it up from where it broke down.

The next question: are those AI-led experiences driving positive outcomes? According to a new blog post by Jodi Reuven, Senior Product Marketing Manager for the Zoom Contact Center, that answer is a resounding yes.

“Since we began offering this technology to customers via the Zoom website, we’ve experienced a 97 percent self-service resolution rate,” she noted. “97 out of 100 customer conversations with our intelligent virtual agent are resolved successfully without involving a live agent.

“This success has driven broader improvements: our CSAT has increased by 28 percent, query no-match rates have lowered by 35 percent, and average handling times have decreased.

Interestingly, Zoom had leveraged a third-party virtual agent before launching and implementing its own offering.

Yet, as the Zoom boom took the company’s growth into a new stratosphere, it recognized the need to implement a more scalable solution.

Ideally, that solution would handle millions of customer requests each month and do so reliably.  

As the enterprise communications giant weighed up this need, it also kickstarted its charge in the contact center space, releasing a CCaaS solution that has flourished ever since.

Cue the light bulb moment, as Zoom realized the opportunity to blend this CCaaS solution and its broader communications platform with a native conversational AI offering.

From that moment, Zoom swooped for the conversational AI provider Solvvy and never looked back.  

Since then, Solvvy has transformed into the Zoom Virtual Agent and helped transform the company’s own customer service operations alongside many of its customers’.

The New-Look Zoom Virtual Agent

Available across voice and digital channels, the Zoom Virtual Agent is much more advanced than the product once known as Solvvy.

For starters, it now offers more out-of-the-box integrations, SDK support, and contextual live agent handoffs.

There’s also native reporting that provides deep insights into customer intents, virtual agent performance, and – critically – the financial impact of AI-driven automation.

Yet, perhaps most notable is the solution’s use of generative AI. With this, the Zoom Virtual Agent quickly learns from the CRM and knowledge base, understands customer needs, and responds with tailored answers.

Late last year, Zoom leaned further into the capability to better leverage data from the CRM and broader CX stack, heightening the level of personalization it provides.

Within its own operations, that change has proven hugely significant, with Zoom now able to pull more in-depth data to broaden the virtual agent’s capabilities.

For example, when responding to queries, the enterprise communications giant’s virtual agent can now reflect on the customer’s specific plan or package.   

That one simple change has had a massive impact. As Yuliya Pierce, Digital Customer Experience Leader at Zoom, noted:

“The recent personalization that we introduced is driving positive results. Since we launched it, our CSAT has increased by over ten percent in just a few months!”

Another recent addition is the ability of the Zoom Virtual Agent to recognize multiple intents within a single query and then guide the customer through them one at a time.

Moreover, the solution recognizes when a customer’s intent changes midway through a conversation. It then pulls the conversation back on track and reassures the customer that they can answer the second query afterward.

That ability to gauge intent, intuitively guide customers, and offer reassurance is significant in bolstering self-service resolution rates and containment.

While that new feature is not on show, the following video highlights several more capabilities available via the Zoom Virtual Agent.

More on Zoom’s Own Virtual Agent Deployment

Since replacing its previous chatbot with its own virtual agent, Zoom’s internal customer satisfaction (CSAT) scores have surged, increasing by 28 percent.

Additionally, the average handling time (AHT) of AI-led conversations has decreased – alongside query no-match rates. Indeed, the latter metric dropped by 35 percent.

Pierce credits the solution’s robust reporting tools. “I really like being able to review the queries and then filter by specific queries to identify what customers are asking about,” she noted.

“I can see analytics around how often we match to an intent or if no match exists. This information shows where I need to adjust or add intents, which is what is helping us bring down our no-query match numbers so quickly.”

Pierce’s comment here underlines Zoom’s more modular approach to conversational AI.

Its support team doesn’t just plug the virtual agent into various knowledge sources and let it run wild. Instead, they go to a more granular level.

The team can consider whether a generative-based approach is best for each contact reason. For specific conversations, a more rigid decision-tree structure may be better placed.

That ability to take a more modular approach to conversation automation and blend generative and rule-based flows is critical to cautious, customer-centric implementation.   

Learn More About the Zoom Virtual Agent

As the Zoom Contact Center keeps winning new customers, expect more companies to consider the vendor’s virtual agent offering.

Yet, the solution is much more than a natural add-on for Zoom Contact Center customers.  

By drinking its own Cool Aid, Zoom demonstrates that it’s a vendor that not only delivers a visionary solution but also delivers tangible results.

Eager to achieve similar outcomes with a virtual agent? Visit: zoom.com/en/products/virtual-agent/

Agent AssistAI AgentsArtificial IntelligenceCCaaSChatbotsGenerative AI

Brands mentioned in this article.

Featured

Share This Post