AI in the Contact Centre: Efficiency Is Only the Beginning

Artificial intelligence is a major talking point across almost every industry, and contact centres are no exception. For me, the opportunity is in using AI where it can make people more effective, whilst protecting the human qualities that good customer service depends on.

We’re still experiencing a push-pull dynamic. Contact centres are being asked to manage costs whilst customers continue to expect knowledgeable, responsive and personal service, especially when handling complex conversations.

AI certainly has a role in helping us work more efficiently, and we're already seeing that in areas such as quality assurance, call summarisation and advisor guidance.

However, I believe efficiency is only the starting point.

If AI reduces the administration after a call, what does the advisor do with that extra time? If a team leader can review quality more efficiently, could they spend more time coaching their people?

And if technology helps us spot a problem sooner, how quickly can we use that insight to improve the next customer interaction?

Those are the opportunities I find most interesting.

How Does the Use of AI Remove Customer Friction?

We've probably all experienced a customer service conversation where we seem to know more about the issue than the person we're speaking to.

It immediately undermines confidence.

For an advisor, that situation is frustrating too. You can have the right skills and genuinely want to help, but you still need accurate information in front of you.

Using platforms such as MaxContact, we're able to present guidance within the live advisor environment and update it as circumstances change. This is enabled by dynamic scripting that enhances the customer experience.

For example, if there is a temporary service issue affecting a client's location, we can add that information to the script. This means the next advisor taking a relevant call has something useful and current to share with the customer.

It gives advisors confidence and helps us provide greater consistency across the team.

The script doesn't have to dictate the conversation word for word. It gives the advisor the information they need, whilst their listening skills and judgement shape how they use it.

How Can AI Give Advisors More Time to Focus on Customers?

Another area we're exploring is automated call summarisation.

Some of the conversations our teams handle are lengthy and detailed. An advisor supporting someone looking for care for a family member, for example, may need to understand personal circumstances, care requirements and several possible next steps.

Recording that properly after the conversation takes time.

We're looking at MaxContact functionality that can listen to the conversation and produce a structured summary for the advisor to check and use. The advisor remains responsible for making sure the information is accurate, but the amount of manual note-taking could be reduced significantly.

This is where I think we need to look beyond the immediate efficiency saving.

The benefits go beyond how many minutes we've removed from wrap time.

Instead, we’re questioning whether that extra capacity allows us to answer more enquiries and provide a better overall service.

For example, could we use this additional time to do more follow-ups? Or could we carry out more proactive after-sales contact?

Saving time only creates value when you make a conscious decision about where to reinvest it.

What is the Role of AI in Improving Contact Centre Quality?

Quality assurance is another area where I see a clear opportunity.

Traditional QA relies heavily on sampling. Team leaders listen to selected calls and assess them against an agreed framework.

That gives you useful insight, but your view is inevitably limited by how many conversations you have time to review.

On the account I support, manually listening to and scoring a detailed call can take up to 30 minutes. Using Spokn AI for advanced speech analytics, we can complete an initial assessment against a defined scorecard in around five minutes before reviewing the findings ourselves.

That means we can improve both the speed and quality of analysis.

Rather than relying solely on a small monthly sample, we can carry out targeted audits after a process change and see whether the new guidance is being followed. We can also look for recurring themes and identify where further support might be needed.

The quality framework still needs careful design and tailoring. A sales-led account team may want to understand objection handling, whilst an account supporting families looking for care might focus more closely on listening, accuracy and appropriate guidance.

We have to test those scorecards, compare the AI output against our own assessments and refine them when something isn't being interpreted as intended.

It’s important to do this upfront work. AI can analyse what you ask it to analyse, but you still need a support team who understand what great customer service looks like.

When AI Reveals a Process Problem

One of the most useful things we've learned is that better quality insight can tell you something about the process, as well as the advisor.

We recently introduced a requirement to collect some additional customer information linked to a client's marketing activity. We communicated the change and then used Spokn AI to check how consistently it was being applied.

The results didn't look right.

When we investigated, we found that the guidance had been interpreted differently across the team. Some advisors were using the way a customer contacted us to decide whether the additional information was needed, even though the customer could have encountered the original marketing and then found the contact details elsewhere.

Once we understood what was happening, we could clarify the issue and separate individual performance from a wider process problem.

Used in this way, AI can prevent teams from falling into the quality management trap of only analysing issues at advisor level. With wider insight, you might discover that the real issue sits somewhere else. Perhaps the instruction wasn't clear enough, a process created ambiguity or people were working from different interpretations.

Better visibility helps you address the right problem.

AI Should Give More Time to Lead

Improving team leader capacity is one of the fundamental benefits of using AI in the contact centre.

Listening to calls, finding examples and completing scorecards can keep team leaders at their desks for long periods.

This work is necessary, but some can be supported by technology.

If AI-assisted call monitoring gives a team leader several hours back, I'd rather see that time spent with the team. More one-to-ones, better coaching and greater visibility of what people are experiencing on the floor all have value.

Feedback can become more timely too.

Instead of waiting for a monthly review and giving someone several areas to work on at once, regular AI-assisted checks can highlight individual development points throughout the month.

By the time the advisor sits down with their team leader, they already understand what they've been working on.

The conversation can then move towards progress and development.

That's a much more interesting use of AI than simply completing the same QA process faster.

How to Turn Time Saved into Better Customer Experience

The same principle applies at an operational level.

If advisors spend less time completing administration and team leaders spend less time manually reviewing calls, the contact centre has additional capacity.

So what happens next?

For one team, that might mean more proactive after-sales service. For another, it could mean following up on important customer interactions or giving advisors more time for cases that genuinely require it.

This is particularly relevant when the contact itself carries emotional weight.

Some of our teams support people looking for care for themselves or a family member.

A caller may be worried, unsure where to begin or dealing with a situation they've never faced before.

I wouldn't want technology pushing those conversations towards an arbitrary time target.

I'd rather use it around the interaction, helping the advisor find information, reducing their administrative workload and helping their team leader understand where support is needed.

That allows efficiency and empathy to work together.

Start with a Problem, Rather than a Wish List

Contact centres need to adapt to today’s customer needs. But there's more AI functionality available than most contact centres could sensibly introduce at once.

My approach would be to start small and find a genuine problem.

Maybe advisors are spending too much time writing notes. Perhaps quality monitoring gives you too narrow a view of performance. Or important information isn't reaching the team quickly enough.

Test the technology against that problem, involve the people who will use it and understand what it changes for the customer. Then build from what you learn.

It's also worth remembering that introducing AI is a change programme. People need to understand why a new tool is being introduced and how it affects their role. You need to give them time to get comfortable with it and listen to their feedback.

We're still learning ourselves. We're exploring what platforms such as MaxContact and Spokn AI can do, testing applications and looking closely at where they can add value.

But the measure of progress shouldn't be how much AI you've deployed across your contact centre. It should be whether your customers are receiving better service, your advisors feel better supported and your team leaders have better insight into what is happening.

AI can give contact centres time and insight back. What we choose to do with this additional time is where progress can be made.

Connect with me if you'd like to chat through AI implementation or get in touch if your brand is exploring the benefits of outsourced customer service.