The most visible use case for AI in telecoms is enhanced customer service. Leading telecom companies in the U.S. such as AT&T, Comcast, and Verizon leverage artificial intelligence in a wide array of processes. The long list includes automated chatbots, personalized offers, and efficiently streamlined customer service processes.
With some exceptions, AI-powered customer service solutions can be divided into two categories:
- Customer service communication
- Customer engagement and personalized user experience
Solving (or improving, at least) each of these problems presents potential savings and increased efficiency for the company.
AI-powered customer service communication
To solve customer’s problems at a scale unfathomable for human agents, the AI algorithms empowering customer communication must process a massive amount of data and interactions. In the telecom business, there’s a lot of data with different nature that can be used to train such algorithms.
AI-powered customer service solutions are often represented by a chatbot interface. But that is not always the case. Sometimes, these algorithms also work in the background, helping to make customer service department’s work more cost-efficient. For example, by analysing extensive background data to help a CS agent to identify a customer’s problem and find the correct solution more quickly.
Here are some examples how AI algorithms are benefitting large US telcos in the area of customer service communication:
- Acting as a gateway between customer requests and help centre/live chat.
- Routing customers/customer requests to the proper agent, and routing prospects with buying intent directly to the sales department.
- Analyzing customer requests together with network data to find the solution to customer’s problem more efficiently.
- Identifying “hot leads” from thousands of emails and routing them to the salespeople.
- Letting customers explore or purchase media content by spoken word rather than remote control.
- Entertainment chatbots operating on telecom operators’ native platforms or through the Facebook Messenger platform.
AI as a customer service agent
Telecoms often apply machine learning algorithms to make the customer service process more cost-efficient. This kind of AI use case is present in AT&T, Spectrum, CenturyLink, and many other well-known telcos.
The AI-powered Ask Spectrum virtual assistant helps customers with troubleshooting, account information or general questions about Spectrum services. The customer inquiries managed by the assistant range from identifying service outages to ordering paid content services. The assistant can either provide users with helpful tips and links to the Help Centre or in case of more complex requests, refer them to Live Chat representatives. As a result, some of the work is loaded off the CS team’s shoulders and they’re left to deal with more demanding cases.
There are several more good examples of AI in customer service in the telecom business.
In 2016, Centrylink implemented their AI-driven assistant named Angie. According to the Harvard Business Review, Angie handles an estimated 30,000 emails each month and analyzes the responses to identify “hot leads” that are then routed to a relevant sales department. The initial pilot showed that Angie could correctly interpret 99% of emails that were processed while 1% were forwarded to human agents. As a result, the company’s sales representatives can save a significant amount of time spent on outreach and follow-ups.
AT&T, the world’s largest telecommunications company, leverages AI to process all “online chat interactions.” In December 2016, AT&T rolled out Atticus, the entertainment chatbot that communicated with users via the Facebook Messenger platform.
In April 2017, Vodafone released its new chatbot TOBi that can assist customers via live chat on the Vodafone UK website. Using a combination of AI and predefined rules, TOBi simulates human conversation and responds to customer inquiries ranging from troubleshooting, order tracking, and usage. Recently, TOBi also acquired the capacity to assist users with the purchase of SIM-only plans. The company is constantly looking for new add-ons to their chatbot.
At MindTitan, we also see a potential for utilizing the location layer and network analytics to improve customer service. For example, if a customer from a specific location reaches out with a problem, the algorithms can check network analytics to identify potential issues or shortages in the area. As a result, the customer service rep helping to troubleshoot will have additional knowledge of what the correct solution might be.