From the Friction Era to the Flow Era – some thoughts following the Call Centre Helper webinar.
I had the pleasure of joining a recent Call Centre Helper webinar to discuss a question that's becoming increasingly common across the contact centre industry: Should voice AI agents replace your IVR?
My answer? It depends.
We had some great presentations and discussion, including an interesting point from Lauren that really stuck with me. We're moving from the Friction Era to the Flow Era of customer contact.
I think that's a great way to describe where the industry is heading.
For years, we've designed customer contact around the limitations of our systems. We've asked customers to navigate menus, choose departments, repeat information and follow processes that make sense to the organisation, but not necessarily to the customer.
Now, with advances in conversational AI, we have an opportunity to change that.
But there's a catch. Just because we can replace an IVR with an AI voice agent, does that mean we should?
From friction to flow
Think about a typical customer calling a contact centre today.
They navigate an IVR menu, select an option that hopefully matches their enquiry, wait in a queue and eventually explain what they need to an adviser.
If they've reached the wrong department, they might get transferred and have to explain everything again.
We've all experienced it.
Now imagine the same customer simply saying, "I need to change my appointment."
An AI voice agent understands the request, checks the customer's identity, looks up the appointment, offers alternative times and completes the change.
No menus. No transfers. No unnecessary waiting.
That's a great example of moving from friction to flow.
But here's where I think we need to be careful.
Putting a shiny new voice AI agent in front of a broken customer journey doesn't fix the journey.
If your systems don't talk to each other, your customer data is unreliable or your processes are unnecessarily complicated, AI won't magically solve those problems.
You might just end up automating the frustration.
Before you automate an enquiry, ask why it exists
This was one of the main points I wanted to get across during my presentation.
In my experience, organisations are often very good at measuring contact volumes, but not always as good at understanding why customers need to contact them in the first place.
Take a customer calling to ask for an update on their application.
It's probably a relatively simple enquiry for an AI voice agent to handle.
But why is the customer calling?
Could we have sent them a proactive update? Could they have checked the progress online? Has something gone wrong elsewhere in the process?
If we can remove the need for that contact altogether, surely that's a better outcome?
Before investing in voice AI, I'd start by looking at:
Your top contact reasons. What are customers actually contacting you about?
Repeat contacts. How many customers are coming back because their enquiry wasn't resolved the first time?
Failure demand. Which contacts are being created by something going wrong in your processes?
Customer effort. Where are we making customers work harder than they should?
Opportunities for automation. Which enquiries could genuinely be resolved more quickly and easily using AI?
Get this right and you'll have a much better idea of where AI can add real value.
Not every customer enquiry should be handled by AI
I'm a big supporter of using AI where it makes sense.
Appointment changes, straightforward account enquiries and other predictable, low-risk tasks can be great starting points.
But what about a customer who's lost their job, is struggling financially and believes they've been charged incorrectly?
That's a very different conversation.
There could be vulnerability, regulatory obligations, several interconnected issues and a real need for empathy and human judgement.
In these situations, I'd be looking at how AI could support the adviser, rather than replace them.
For example, AI could help retrieve relevant information, summarise the customer's history or suggest the next best action.
The important thing is recognising that the right outcome isn't always the one that involves the least human interaction.
Your frontline teams need to be involved from day one
One of my favourite takeaways from the webinar actually came from the audience.
When asked for their top tip, the message was clear: involve your frontline teams from the very start.
I couldn't agree more.
Your advisers deal with the reality of your customer journeys every single day.
They know which processes don't work, where customers get confused, which systems create unnecessary work and what happens when something doesn't follow the expected process.
Yet I've seen projects where the people who actually use the systems are barely involved until it's time for testing or training.
That makes no sense to me.
If you're redesigning your IVR, improving a customer journey or introducing an AI voice agent, involve your frontline teams when you're defining the problem, not just when you're testing the solution.
They'll often spot things that a project team or technology supplier might miss.
And if you're introducing AI, they're also the people who will need to manage the enquiries that AI can't resolve.
Their role becomes more important, not less.
Launching an AI agent is just the beginning
Another area I covered was what happens after an AI voice agent goes live.
Who owns it?
Who reviews the conversations, identifies incorrect responses, updates the knowledge and decides when changes need to be made?
Is it Operations, IT, CX or a combination of all three?
In my view, this needs to be agreed before launch.
You also need to think carefully about how you're measuring success.
It's very easy to get excited about containment rates and cost savings.
But a call that never reached an adviser isn't necessarily a successfully resolved enquiry.
Did the customer actually get what they needed? Did they have to call back? Was the experience easier than before?
As I said during the webinar:
Containment is a contact centre metric. Resolution is a customer outcome.
We need to measure both, but we shouldn't confuse the two.
So, should voice AI replace your IVR?
For some organisations and some customer journeys, absolutely.
For others, it might make more sense to improve the IVR you already have, introduce AI for specific enquiries or focus first on fixing the underlying processes.
I'd encourage any organisation considering voice AI to start small.
Choose a clearly defined, high-volume, low-risk enquiry. Make sure your systems, information and processes can support it. Test it with real customer scenarios, including the awkward ones, and involve your frontline teams throughout.
Then expand based on what you learn.
The move from the Friction Era to the Flow Era is an exciting opportunity for the contact centre industry.
But the objective shouldn't simply be to replace old technology with something new.
It should be to make life easier for customers and the people who support them.
And ultimately, that's how we should judge whether an investment in AI has been worthwhile.
At NextGenCX, we help organisations understand their customer contact, identify avoidable demand and find practical ways to improve customer journeys, systems and processes. If you're considering introducing voice AI into your contact centre, we'd be happy to have a conversation.
Steve Hindley
Founder & Lead Consultant, NextGenCX
