Harnessing AI in a contact centre to deliver enhanced CX and reduce overheads

Achieving the balance between delivering high-quality customer experience and maintaining financial stability within your contact centre requires careful navigation. While acknowledging the range of available technologies, it’s essential to recognise that they often come with associated costs.

Similarly, increasing the number of agents can lead to shorter queues and increased customer satisfaction. However, if the cost of hiring additional agents renders your contact centre unprofitable, you’re continuously struggling to justify the return on investment. While it may appear that more agents equate to better service, within this guide we explore the other possible options available to resolve this common headache.

The good news is that the rapid innovation into the area of Artificial Intelligence (AI) has come to the fore in recent years, presenting an ideal solution for your contact centre.

By leveraging AI in a contact centre, you can streamline agent workflows and optimise the customer journey, effectively balancing your expenditures and transforming your contact centre into a profit centre. But, most importantly, significant cost savings can also be realised through successful AI implementation.

From in-app agent coaching to accurate staff forecasting and self-service capabilities, AI functionality aligns seamlessly with the day-to-day operations of your contact centre, and once implemented, businesses experience a substantial reduction in average operational costs.

download the guide as a PDF

The average cost of contact centre operations

To find the average cost of contact centre operations, we’ve taken average calculations from a number of public sources.

These figures will vary from business to business and it’s best to scrutinise your own costs before making important business decisions.

The following are included as an example of a single agent cost, to provide a clear picture of the average cost of contact centre operations.

We can therefore make some estimates by multiplying your contact centre employee count by £32,000. The average contact centre in the UK has 127 employees. That makes the annual average cost of contact centre operations a fraction over £4,000,000. Costs per number of employees are indicated below.

You must also factor in cost per call, accounting for agent effort, availability and monetary cost. At a rate of £5.58 per call and an average of 4,400 calls per month, that’s an extra £24,552 per year.

Omnichannel contact centres

By adding extra communication channels, the average cost per transaction decreases.

The average cost per email is £3.55
The average cost per web chat is £3.05
The average cost per social media is £2.25

Staff turnover

With the average staff turnover rate in the UK at 26%, for £4m you spend on contact centre costs per year, you can expect to pay another £1m due to staff turnover. An alternative to running your own contact centre is outsourcing. Depending on location, costs vary between £1,500,000 and £17,000,000 per year for a 127-employee contact centre.

Why are contact centre overheads so expensive?

Considering all the elements involved in operating a contact centre reveals some quite alarming expenses. There is a stark contrast between an agent’s hourly wage and the overall costs of a contact centre. Overheads are a crucial target for cost-saving endeavours. However, before embarking on cost-cutting measures, it’s important to initially understand why your figures are so high.

Staff Retention

It should come as no surprise that people are the most expensive cost centre in a contact centre. Yet, without them there is no operation at all.

Whilst you could pay agents less, this isn’t a viable option. To provide the best service, you need the best talent.

One of the most concerning costs, however, is the price of retaining (or not retaining staff). With agent retention in contact centres significantly surpassing the national average of 15%, the loss of a staff member entails substantial costs.

Recruitment starts again and is swiftly followed by training. Unfortunately, some of these freshly hired agents never make it to their first day on the phones. For those who do, it typically takes between four and 10 weeks before they achieve proficiency in handling calls and can operate independently.

Once they are up and running, it’s vital that you do everything possible to look after your agents. Failure to do so leads to poor employee satisfaction and the desire to find a new employer.

When contemplating strategies to increase staff retention, consider implementing the following staff initiatives:

The cost of keeping employees happy versus finding and training new ones is heavily weighted towards retaining your existing team. Reducing contact centre overheads is not accomplished by simply cutting costs related to personnel.

Legacy Contracts

Are you still using a phone system you implemented a decade ago, hoping to eventually see a return on your investment? You’re not alone. The reluctance to invest in new technology often stems from the challenge of replacing existing infrastructure.

You may instinctively believe that you have at least one more year before your current system becomes obsolete. Then you’ll start considering a replacement and transition to cloud-based solutions.

However, every moment you delay embracing modern cloud technology, you’re incurring hidden costs for your contact centre.

The lack of productivity caused by insufficient functionality manifests itself in two major ways:

Relying on outdated technology until it becomes obsolete is rarely a good solution, especially when modern alternatives offer improved operational productivity.

Upfront Costs

Historically, installing contact centre hardware required a major upfront payment. Today, thanks to the advent and success of the cloud, the need for hardware is minimal. There will always be a need for a computer and a phone/headset but, other than this, almost all contact centre functionality is software-based and accessible via the cloud.

The combination of lack of hardware and ease of deployment also removes the need for professional services in many cases.

The modern agent interface can be accessed by means of a username and password through a web browser. When in the app/browser, getting started is simple, minimising training and adoption time. The current contact centre display is notably more intuitive compared to its counterpart from a decade ago.

Traditional upfront contact centre costs play a huge part in overheads. Businesses who opt for a cloud contact centre also enable countless efficiencies through AI integration.

Lack of cost optimisation processes

Contact centre overheads are often high costs. While agents spend time firefighting customer issues, IT is overloaded with support tickets, and supervisors are handling escalations, leaving little time left to review costs.

When there are no resources available to scrutinise costs, yet optimisation is necessary, we reach a stalemate.

You have the option to implement some of the following tried and tested solutions:

Hire a business consultant: review agent performance and conduct a gap analysis
Embrace call deflection strategies: features like IVR, ACD, and chatbots
Outsource to self-service: knowledge bases, virtual agents, and auto-onboarding
Implement workforce management: dynamic, AI-powered forecasting based on genuine data

If you lack a defined cost optimisation process, the right technology could be doing the hard work for you. Instead of finding the time and resources to seek cost savings, let software do what it’s designed to do.

The mundane, data-driven tasks that take hours for humans to process and analyse can be completed in seconds with the use of AI in a contact centre which is all part of modern contact centre technology.

Adopting new technologies to deliver a clear ROI

While technology isn’t the only solution for improving contact centre operations and reducing overheads, it undeniably plays a significant role.

The technologies that can perform number crunching and data manipulation, and translate those actions into business outcomes are the ones you’ve been missing out on the most.

These are the technologies with a clear return on investment. For example, if you can demonstrate that you save 50% on agent shift fees, equating to £100,000 per year, a workforce management tool that costs £1,500 per year becomes an obvious weapon in your arsenal.

The same is true for other AI-assisted tools like IVR and chatbots. If these tools, which are now smarter than ever thanks to natural language understanding and machine learning, can deal with 25% of call volume, agents can spend time on more value-adding tasks.

When agents spend quality time resolving customer enquiries, first contact resolution soars and you have a happier base of customers. They stay with you for longer, continue to spend money, and spread a positive sentiment about your business to others.

So why the lack of adoption with AI in a contact centre?

A key factor here is fear of change. More specifically, the fear of the unknown.

What happens if we adopt AI?

How will our customers feel if we use AI behind the scenes?
How will our customers feel if we introduce self-service?
How will our agents feel if we automate some of their tasks?
What if it works and we want to roll out at scale?

There are two things we need to do here:

Learn from existing case studies
Ask our employees and customers for their thoughts

The use of AI in contact centres is widely communicated:

90% of customers are satisfied with automated chatbots
MoneySolver has automated 100% of quality assurance
TruConnect has reduced average handle time by 30 seconds
North Yorkshire Council freed up agents with a 24/7 chatbot for basic enquiries

As for talking to employees, we need look no further than the headline in a blog post by Workday:

Employees are more optimistic than many leaders think

This is backed by data from a Forrester study that suggests:

If both customers and employees are optimistic and comfortable with the use of AI, the focus shifts from “why not” to “how”.

Here’s how you can enhance operations and reduce overheads using AI in a contact centre

Without extravagant unwanted costs
Without losing staff or compromising on quality
Without dedicating time to lengthy implementation

The majority of cloud contact centre platforms are incorporating AI to such an extent that it has become just another feature. Whether through proprietary technology or the integration of established AI models, functionalities within contact centres that previously depended on manual input and user effort are now automated. These include:

Recommended Contact Centre Technologies

Although we advocate prioritising people and processes before technology, there are still valuable tools and features you could be taking advantage of.

AI Insights

If you identify where your teams are losing time, you can start to do something about it. AI Insights introduces a dashboard breaking down key metrics per topic. This way, you can start to understand the types of queries that consume the most agent time. Using ChatGPT technology alongside real-time transcription, AI Insights automatically detects the topic of customer conversations and sorts them into clusters.

For every topic, you can view:

Number of interactions
Average handle time
Average hold time
Average queue time
Impact score

For instance, when a significant portion of your time is spent addressing queries related to new device activation, you can identify this pattern through clear, databacked insights and take proactive measures to minimise the volume of calls to your contact centre regarding this issue.

Workforce Management

As we know people are the most expensive contact centre overheard, it pays (literally) to get your forecasting right.

Leveraging machine learning algorithms and predictive analytics, you can lean on automated workforce management for the following use cases:

Workforce planning
Forecasting and scheduling
Schedule adherence and performance management
Employee engagement

AI-powered algorithms can analyse large volumes of data and make accurate predictions and recommendations for optimising workforce management. This will help to improve your estimations when it comes to staff forecasting.

Best of all, a human only needs to read it, bypassing days of spreadsheet work to reach the same conclusion an AI tool achieves in seconds.

Self-service IVR

Unlike a traditional IVR, which takes customer input and directs their call to a specific agent or department, an AI-assisted self-service IVR can handle many common queries without the need for human intervention.

For example, a customer calls and presses 1 for support. Your AI-assisted IVR can ask for more information about which type of support and take a number of actions:

Assist with the call using natural language understanding
Offer self-service options like knowledge bases and portals
Transfer the caller to the most appropriate agent using skills-based routing

What we’re doing here is creating “digital offload”. Your call centre queues are dissipating and being filtered into self-service channels like live chat and knowledge bases.

We know that 91% of customers are happy to use a self-service knowledge base if one exists, so making them aware of such service is crucial here. Not only will this remedy their current enquiry, but it will also become their first port of call for support next time.

Interactive Virtual Agents

When a customer visits your website, their expectation is to do one of three things:

Buy a product
Find out information about a product
Get support for a product

While making a purchase and accessing information are the primary objectives of most websites, they frequently lack sufficient support functionalities. In many cases, a web chat bubble is the only option.

This, however, relies on agent availability, and wait times can reflect those of a call queue. Some attempts have been made to incorporate basic chatbot functionality such as offering opening times, verifying user accounts and logging tickets. However, an Interactive Virtual Agent eliminates the need for queues.

Powered by machine learning algorithms and seamlessly integrated with company systems, Interactive Virtual Agents gain context and content, enabling customers to engage in meaningful conversations.

Agent Assist

When agents are under pressure to resolve calls in a timely manner, several things happen:

Agents rush, focusing on their average handle time metrics
First contact resolution decreases as the whole problem hasn’t been remedied
Customers become unhappy and are more likely to leave
Opportunities to up-sell and cross-sell get missed

Every time an agent makes a rash decision, it adds to your overheads. Every time an agent misses a sales opportunity, a chance to remedy those rash decisions passes by. If you implement agent assist technology, an in-app prompt is triggered whenever keywords, phrases or emotions are detected. This provides the agent with a reminder preventing them from missing a key part of the call.

For example, if a customer asks for the shipping time for the product they just purchased, agents get an AI assist prompt with the answer for that particular product. This saves both the agent and customer time during the call and ensures the correct information is shared.

Similarly, in cases where a customer experiences a technical issue that your premium product can resolve, agents might not always be informed about new features or may perceive sales-related tasks as beyond their responsibilities.

This is where agent assist proves helpful once more. Identifying a common issue and cross-referencing it with other products in your portfolio, agents receive a prompt suggesting a new product available for the customer to upgrade to. Rather than just logging a fault and enduring hours of waiting for problem resolution, you secure a sale. An added advantage is that you can address issues without waiting until the stage of retrospective quality monitoring, where it may already be too late.

There’s a reduction in overhead costs through automating quality monitoring, along with an opportunity to achieve break-even during a support call.

Quality Monitoring

Manually listening to hours of call recordings is a time-consuming task. With tools like sentiment analysis and keyword recognition, you can record, analyse, and flag every single call in your contact centre without lifting a finger.

This allows you the time to dedicate to training agents when negative calls arise, enabling you to address the root cause rather than engaging in case-by-case firefighting. What’s more, this quality assurance process happens in real time. For instance, when a call becomes heated, such as when a customer raises their voice or uses bad language with an agent, supervisors can receive real-time notifications through the app, email, or even an audible beep.

With AI in a contact centre, systems automate quality monitoring by looking after the following aspects:

How to kick off your AI implementation

The landscape of contact centre operations stands at a crucial turning point, where the convergence of technology and strategy holds the potential to elevate operations and reduce overhead costs.

By harnessing AI-powered tools such as Workforce Management, self-service IVRs, Interactive Virtual Agents, Agent Assist technology, and Quality Monitoring, contact centres can enhance customer experiences while driving significant cost savings.

The good news if you’re already using cloud contact centre technology is that all these features are readily available today. If you’re not in the cloud, transitioning from an on-premises system to the cloud has never been simpler.

Start reducing your contact centre overheads with AI. Contact us to book a call with one of our implementation experts.

Download this useful guide now as a pdf

Achieving the balance between delivering high-quality customer experience and maintaining financial stability within your contact centre requires careful navigation. This handy guide explains how you can achieve exactly this.