AI and automation are rapidly transforming customer experience (CX) and contact centre operations, turning these into differentiators rather than simple customer support. These tools now enable sentiment and call analysis, as well as personalisation and improved operational efficiency. Customers are, in theory, more satisfied and the business gets the benefits of shorter waiting times and better outcomes.
Or do they? According to the Puzzel ‘State of Contact Centres 2025’ study, around 61% of leaders now trust AI chatbots to resolve complex queries. This is a lot of confidence in a technology that’s delivering mixed results. Yes, it boosts speed and efficiency, but it also introduces risks. When chatbots aren’t managed well, they undermine trust and emotional connection, which may result in negative customer experiences when people want empathy or nuanced support.
Meanwhile, companies are leaping into AI. According to Gartner, 85% of customer service leaders have embraced AI, but customers are stepping back; 77% say they find chatbots frustrating and most people would rather speak to a real person, someone who understands complex issues and can offer real support. The challenge, says the Harvard Business Review, is to strike a balance between the psychological needs of the customer and the demands of the modern call centre.
Even when call centres were manned exclusively by people, with only a few doddering self-service options, customers still slammed down the phone in frustration. In 2009, a study focused on call centre satisfaction in South Africa found that people were, “less than 30% happy with call centre service delivery”.
Instead of waiting for customers to reach their breaking point, automated monitoring predicts issues and triggers outreach before frustration kicks in.
Saša Slankamenac, Dariel
Industry resources and annual studies from firms like Zendesk and McKinsey, among others, show that frustration with call centre experiences remain stubbornly persistent despite the technology improving incrementally over time. Since the pandemic, this has worsened considerably, as people want more solutions, faster, and with more accurate personalisation. McKinsey found that 61% of leaders have reported increased call volumes since 2020 and 71% of customers expect personalised interactions. Customers still prefer to make a call, even for non-urgent issues.
There is also the fact of the high employee turnover in the contact centre industry. The average rate is 30% to 45%, with most agents lasting just over 13 months in their roles. High stress levels, low salaries, and little prospect of career growth contribute to the situation which, according to an Insignia Resource study, costs companies anything between $10 000 to $20 000 per employee to replace. Numerous studies have shown that contact centre agents tend to get burned out very quickly, resulting in detached and disinterested agents who do little to bolster the company’s reputation or build customer engagement.
Alongside automation and workflow management, AI will also become increasingly used to link smart devices to contact centres for proactive monitoring and support of infrastructure.
Moving back to the psychology of interactions between humans and machines, a group of universities undertook a study to examine human reactions to technology. They found that even when people were speaking to chatbots in exactly the same way as they were to humans, their satisfaction dropped by 8.5%. The perception that they were handed off to a machine, among other factors, immediately had a negative effect on the interaction. But when they were told the chatbot was available to them any time of day or night, and that it could provide them with immediate answers to their problems, they were 37% more satisfied with the technology.
AI is also changing how agents engage with customers. Yes, there are the same concerns around job safety as agents worry that chatbots are about to step in and take over, but the reality is very different. For companies that want to continue to employ people, AI is a tool that empowers agents and potentially reduces their emotional load.
Martie de Beer, Contact Centre-as-a-Service executive at Telviva, has found that the area where the company is seeing the most growth is in the components that simplify an agent’s life. “Knowledge base integrations, easier access to transcription services and other key aspects of the role are centralised with AI,” she says. Agents have to toggle through a lot of screens and channels, from WhatsApp to emails to documents, so if AI can consolidate these fragments, then it can have a transformative impact on how they do their jobs.
But sticking AI into a system that remains disconnected and locked in silos won’t empower anyone. The technology has to work within a centralised system that gives agents the insights they need so they can talk to people from a point of understanding. People are calling because they need resolution to questions that an AI can’t provide, so AI’s role must be more attuned to supporting the people behind the desk.
The technology assists in streamlining processes so human agents can focus on more complex, high-value remediations that require more creativity and technical prowess.
Fikile Sibiya, e4
“Intercape had so many queries using text and phone-based communication to get answers to queries that ranged from, ‘Can I bring my chicken on the bus?’, to ‘How big is my bag?’,” says De Beer. “Agents were fielding these constantly, so we analysed their voice calls and automated the system on WhatsApp and webchat with a vast knowledge base designed to handle these mundane day-today questions.”
This integration was just in a small call centre with 30 people, and the agents had 50 000 fewer chats to manage, achieving an impressive total of 190 000 chats in just three months. When this level of automation is applied to a larger contact centre, it’s easy to see how it could instantly alleviate significant stressors for agents and company alike. “Agents can also use real-time copilots to get helpful tips, suggested next actions or coaching tips mid-conversation with a customer,” says Saša Slankamenac, architect in office of the CTO, Dariel. “It reduces their cognitive overload and frees them up to focus on emotional intelligence rather than on endless system navigation. The result is that interactions feel warmer and more human.”
When an agent can anticipate a customer’s needs, the interaction improves. People are less frustrated, emotions are less negative, and it is easier for agents to find resolution to their problems.
“AI offers the contact centre the ability to have hyper-personalised interactions with customers based on historic interactions, giving them the space to adjust the tone of their engagements based on data,” says Fikile Sibiya, CIO, e4. “The technology assists in streamlining processes so human agents can focus on more complex, high-value remediations that require more creativity and technical prowess.”
AI takes on the low-level tasks, which is precisely where its value lies. Or does it? Increasingly, AI is being used to directly address negative customer perceptions by making interactions feel more natural and empathetic, effectively bringing humanity into the contact centre. These tools use real-time language processing and sentiment analysis to interpret customer emotions so the system can then tailor responses or escalate queries to human agents when empathy is required.
Currently, there are several platforms that offer AI in their cloud and localised contact centre services. Amazon Connect, part of the AWS cloud-native contact centre platform, can now be deployed with in-region latency and data residency thanks to local AWS regions. Cisco Webex Contact Centre has been available in South Africa for a while, running from a new local point of presence hosted on AWS; and Genesys partnered with Accenture and MTN in 2024 to deliver a Cloud-native Contact Centre as-a-Service (CCaaS) solution to African enterprises.
Machine learning models are also adept at analysing behaviour, past interactions, sentiment history and service gaps to calculate the risk of a customer dropping off a call or complaining. High-risk customers get immediate attention, or they’re swiftly routed to a senior agent.
De Beer, from Telviva, believes that one of the important changes to the technology and the business is going to be the prioritisation of unification. “Companies have disparate communication platforms and so I believe the next 12 to 24 months will see an amalgamation of multiple channels into a unified communications platform that’s supported by AI so agents have the right information on demand.
“We’re probably going to see continuous growth in the layers of integration with AI tools used for intent-based routing, and journey orchestration is very much going to remain a priority for contact centres wanting to improve customer retention and service while equally reducing stress on agents,” she adds. “This will be complemented by the ability to use LLMs on the text-based side of contact centre management.”
Alongside automation and workflow management, AI will also become increasingly used to link smart devices to contact centres for proactive monitoring and support of infrastructure. Sibaya says conversational analytics and predictive analytics will grow even more prevalent as companies use them to forecast demand, manage elasticity, and target and personalise engagement.
It’s unlikely that AI will take away jobs. Rather, it looks like the technology will change the way agents work and give them more opportunities for growth and improved job satisfaction. And perhaps, as agents feel less bombarded and stressed, customers will have better experiences, and AI will smooth over the rough edges that have made people slam down phones and lose their minds since the first contact centre in the 1960s.
* Article first published on www.itweb.co.za
Share