Telecom operators around the world are laser-focused on upgrading their wireless networks from 4G to 5G so they can deliver faster network speeds with less delay and accommodate all types of new applications, from cloud gaming to streaming video. However, the move from 4G to 5G creates more demands on the network and adds many more layers of complexity. For example, telecom operators now are not just providing wireless connectivity, they also offer TV and home networking services, and even assist customers with device setup.  All of this means that customer care programs must be more sophisticated so customer issues can be diagnosed and solved.  And the only way to handle that additional complexity is through automation and artificial intelligence (AI).

Dash Research’s recent report on CX in the Telecommunications Industry found that many telecom providers are using automation (driven by machine learning [ML] or other AI technologies) to handle more complex tasks so that network engineers can focus on network issues. For example, AI-driven automated network management tools can be used to handle administrative tasks, such as order processing and invoicing, and can also be used for more complex processes, such as customer onboarding and offboarding.

“The complexity of networks is growing,” said Thomas Kinnman, head of AI & automation development with Ericsson’s managed services business area, adding that this also means that the network is producing a lot more data about subscribers and all that data needs to be managed.  That is why AI becomes such an important part of CX. “Everything in AI is about data. You have to have a good data management strategy as a telco,” Kinnman said.

But some telcos are better at handling AI than others. For example, Ericsson emphasizes the importance of having a data management strategy throughout all parts of the network, as opposed to siloed solutions that do not communicate across the network. “You see a lot of niche players,” Kinnman said, referring to companies that offer AI-based solutions, but only for one area, such as in the radio access network (RAN) or customer care department. But those niche solutions do not usually communicate with the rest of the network. “You really need an end-to-end solution to leverage the data and understand how this is working,” he said.

One example of how important it is for AI-based solutions to communicate across the entire organization is that for many operators the only way they learn about problems within their network is when a customer notifies them that something is happening. Kinnman said that if an operator had different AI systems that communicated throughout the network, predictive analytics would alert them to problems in the network before it impacts customers. “It’s very important to have automated processes to detect even the small things,” he said.

In the Ericsson Mobility Report June 2021, the company pinpoints one technology that is helping telecom providers optimize CX as they incorporate 5G into their networks.

Reinforcement learning (RL) is a branch of ML and AI that enables the network to learn from experiences and constantly improve CX. For example, rather than only responding to manually programmed behaviors, the AI agents are able to learn and even optimize complex processes autonomously without intervention from a human agent. RL is becoming more important with 5G because 5G networks are so complex that it is almost impossible to manually create rules for every use case. However, by using RL, the AI agents are constantly adding to their experience so they can determine the optimal policy for any scenario they encounter.

But the move from 4G to 5G is not the only reason telcos are becoming more interested in using AI (and more advanced technologies like RL) to handle customer care. In the aftermath of the COVID-19 pandemic, telcos are finding that customers are increasingly turning to online customer care, as opposed to using operator brick-and-mortar stores to take care of their customer service complaints. Kinnman said that this trend is making operators place more emphasis on their digital channels and on AI. “Some operators are better at handling this and we know that a lot of our AI-products got a boost from this,” he said. But perhaps the biggest reason for telcos to invest in AI for their customer care is that this technology can help them provide better service to their customers, which often leads to upsell potential. “Telcos are starting to look at how AI and automation can get them more revenues,” Kinnman said. “Customers that are very satisfied are easy to upsell. And that’s making AI more interesting.”