Customer service automation has evolved significantly over the last few years, and what began with simple self-service journeys and basic chatbots is now progressing towards AI-powered virtual agents capable of handling increasingly complex customer interactions.
However, the goal isn’t to eliminate human advisors altogether, Instead, contact centres are learning how to combine AI and human expertise more effectively, ensuring customers receive the right support at the right time.
To find out more, we asked Martin Taylor, Co-Founder and Deputy CEO at Content Guru, to explain how AI-powered virtual agents are steadily moving up the customer service automation pyramid and what this means for the future of customer experience.
Video: Virtual Agent Escalations: Think About the Challenge as a Pyramid
Watch the video below to hear Martin explore how to minimize unnecessary escalations from virtual agents while improving the overall customer journey, focusing on why contact centres should think about the challenge as a pyramid:
With thanks to Martin Taylor, Co-Founder and Deputy CEO at Content Guru, for contributing to this video.
This video was originally published in our article ‘Are Your Virtual Agents Escalating Far Too Many Queries?’
What Is the Escalation Pyramid?
One useful way to think about virtual agent adoption is as a pyramid, as Martin explained:
“If you think about the challenge as a pyramid, at the bottom you have the high-volume, low-complexity interactions that traditionally clog up contact centres – those are already being heavily automated.
For example, at our customer UK Power Networks, 94% of all power outage communications are now handled fully automatically.”
At the bottom sit the high-volume, low-complexity enquiries that have traditionally consumed significant contact centre resources, and these are the interactions that virtual agents are already handling successfully today.
Examples include:
- Account updates
- Status enquiries
- Password resets
- Service notifications
- Frequently asked questions
For many contact centres, virtual agents are already managing the majority of these interactions without human involvement.
How Virtual Agents Are Taking on More Complex Customer Interactions
While virtual agents initially focused on simple enquiries, advancements in AI are allowing them to support increasingly complex customer journeys, and modern virtual agents can now:
- Understand customer intent more accurately
- Manage multi-step conversations
- Gather contextual information
- Guide customers through more complex processes
- Resolve a wider range of issues independently
“What the industry is working through now is progressively moving further up that pyramid, tackling the next lowest hanging fruit. Agentic AI or virtual agents are becoming capable of handling increasingly complex use cases.
Of course, some interactions will still need to be escalated to a human, whether that’s because the customer reaches the guardrails of the AI’s permissions or because they simply ask to speak to a human.”
As a result, contact centres are steadily expanding the types of interactions virtual agents can successfully resolve without requiring human intervention.
Why Escalating to a Human Agent Doesn’t Mean Failure
One of the biggest misconceptions about virtual agents is that every escalation represents a failed interaction, Martin continued:
“Importantly, escalation doesn’t mean failure. Even when a conversation is handed over, the virtual agent has already removed a large amount of repetitive work for the human advisor, collecting customer details such as the reason for contacting and verifying information when necessary. The human agent receives the full interaction history and can immediately focus on solving the problem rather than restarting the conversation from scratch.”
In reality, successful customer journeys often involve a combination of virtual and human support.
Even when a customer is transferred to a human advisor, the virtual agent can still provide significant value beforehand by:
- Capturing the reason for contact
- Authenticating the customer
- Collecting key information
- Identifying the nature of the issue
- Creating a complete interaction history
This means human agents don’t need to start the conversation from scratch, and instead, they can immediately focus on resolving the customer’s issue, reducing effort for both customers and advisors.
How Leading Contact Centres Are Using Virtual Agents at Scale
Many contact centres are already seeing significant results from virtual agent deployments, and Martin highlighted the example of the Driver and Vehicle Licensing Agency (DVLA), which handles hundreds of thousands of virtual agent conversations every month:
“A good example is the Driver and Vehicle Licensing Agency, DVLA, which handles more than 250,000 AI webchats every month, rising to 350,000 during peak periods. Over 30% of those interactions are resolved entirely through AI self-service.”
This demonstrates how virtual agents can successfully resolve a substantial proportion of customer enquiries while ensuring human advisors remain available for more complex situations.
Reducing Escalations Through Continuous Improvement
The most successful virtual agent programmes do not stand still; instead, contact centres continuously analyse where escalations occur and look for opportunities to improve customer journeys.
This typically involves:
- Identifying common escalation points
- Understanding why customers request human support
- Refining virtual agent journeys
- Expanding the range of issues virtual agents can resolve
Over time, these improvements allow virtual agents to handle more interactions independently while maintaining a positive customer experience.
“The key to reducing escalations is continuous iteration. Organizations refine customer journeys over time, identifying where escalations happen and gradually designing them to be captured by the virtual agent.”
Keeping Human Agents Focused on High-Value Conversations
As virtual agents take responsibility for more routine interactions, human advisors can focus on the situations where they deliver the greatest value, such as:
- Complex enquiries
- Urgent issues
- Emotionally sensitive conversations
- Vulnerable customer support
- High-value customer interactions
The goal is not to replace human agents, but to ensure their expertise is focused where it has the greatest impact, as Martin concludes:
“As AI capabilities mature, automation moves further up the pyramid. Ultimately, human advisors become focused on the interactions where they add the most value – complex, urgent, emotional, or sensitive conversations, alongside supporting vulnerable or high-value customers.
The end goal is not replacing humans but making every interaction more efficient and customer-centric.”
The future is about creating a partnership between virtual agents and human advisors, where technology removes repetitive work and people focus on solving the problems that require empathy, judgement, and experience.
If you are looking for more great insights from the experts, check out these next:
- 3 Tips to Reduce Escalations From Virtual Agents
- 3 Ways AI Is Transforming Customer Preferences and Personalization
- How to Tailor Interactions to Individual Customer Preferences
Author: Robyn Coppell
Reviewed by: Jo Robinson
Published On: 31st Jul 2026
Read more about - Video, Agentic AI, Artificial Intelligence (AI), Content Guru, Conversational AI, Customer Experience (CX), Handling Customers, Martin Taylor, Service Strategy, Technology Enablement Strategy, Technology Roadmap, Videos
