
UK ISP Virgin Media (O2) have this morning followed yesterday’s news from Liberty Global by announcing the expected launch of a new AI voice agent, which will initially be deployed to support selected broadband fault calls (i.e. helping customers get routine issues resolved more quickly) and will later “gradually” be expanded to cater for other issues.
At present the “routine broadband fault calls” that Virgin Media seem to be targeting are only said to represent a “small proportion of total call volumes“, while VMO2 promised that human agents would remain available to support customers with more complex issues. As per yesterday’s news, this is all being built using Sierra’s conversational AI technology.
The launch is said to build on VMO2’s “ongoing customer service transformation“, including its Institute of Customer Service (ICS) partnership, the introduction of a specialist UK-based team (more than 500 strong) for complex customer support needs, Lumi AI tools for advisors, and “wider investment in improving customer experience“.
Advertisement
Suffice to say that the announcement goes to great lengths to say that this is all about improving support for consumers, although it does occur at the same time as the company are known to be consider a major cost cutting drive and the related potential for further job cuts. How much impact all of this may have on their future performance is yet to be known.
However, VMO2 points to how Ofcom’s official data shows they’ve delivered “sustained reductions in complaints from … customers over the past 18 months“, as well as an increase in the proportion of complaints resolved within 24 hours.
Alan Stott, COO at Virgin Media O2, said:
“Customers want support that is quick, simple and effective, and our people want the right tools to help them deliver brilliant service. Launching this service is another step forward in our plan to use technology responsibly to improve the customer experience.
This is about supporting our people, not replacing them. By using AI to help with routine journeys, we can give customers faster support where it makes sense, while freeing up our advisors to focus on the complex, sensitive and human moments where their expertise and experience make the biggest difference.
We’ve made real progress improving customer service, from investing in specialist teams and accredited training to rolling out AI tools like Lumi that empower our advisors. But we know there is more to do, and we’ll continue testing, learning and investing so we can deliver even better service for our customers.”
Much as we said yesterday, the big challenge here is that consumer sentiment toward the use of AI chatbots and voice agents tends to be quite mixed, with many viewing it as being more of a negative (i.e. a way of reducing the number of actual humans that are available to provide support) – often regardless of whether or not they help to resolve problems more quickly.
On the other hand, if such systems do end up making it quicker and easier for customers to get their issues resolved and the perception challenge can also be overcome, then this could still end up being a positive change. Time will tell.
Advertisement
Advertisement
The new way of going round in circles via an automated bot. Intially, I’d expect some cop out like a button to press to speak to a human, then once its going for a while and VM THINK they’ve covered scanarios, it will be impossible to speak to a human. Could be wrong, but don’t think i’ll be. It’s all about cutting staff (cost).It’s an American company, anyone who doesn’t generate revenue is seen as “cost” – even though they may be vital in keeping customers….
I tend to find with most of the more sophisticated LLM based AI agents that asking to speak to an advisor two or three times usually gets you though to a human, albeit often with a wait time before you can actually speak to one.
Having dealt with various human support agents who go through the “have you rebooted your router” stuff, I think a sufficiently trained LLM will do a superior job at triaging issues and getting the right people’s attention to get any problem fixed.
“You’re absolutely right, and thankyou for correcting me on that”
Virgin Media broadband support is one of the worst I have ever dealt with. A useless chatbot wastes your time with completely irrelevant questions, and after waiting forever to finally reach a human, all they say is: “We’re having system issues today, call back tomorrow.” They don’t care at all that people working from home depend on a stable internet connection.
Their modem/router is another nightmare. If the Virgin Media Connect app acts up, you can’t even change basic settings. It’s completely broken—I had to factory reset the entire device just to get it working again.
I would never recommend Virgin Media to anyone. I will be switching to another provider the moment my contract allows.
Reality strikes here…. ISP’s like Virgin know when someones broadband is down so they want to make a system that eventually discourages contact at all and make the customer wait or do the leg work to fix issues.
Pretty sure back in the day when the like of Bulldog and AOL existed this was pretty much a similar process other than customer said internet not working they said connected to them so tough sh**.
All the Ai does is makes for better automation and better association and recognition of issues.