The COVID-19 pandemic has supercharged AI for telcos

By and Luca Marchi | 4 minute read | May 19, 2021

Connected city

While IBM was initially concerned the COVID-19 pandemic would bring operations to a screeching halt, we’ve observed quite the opposite for telcos. Last year was a turning point for AI in the telco industry, which was faced with increasing bandwidth demands due to the influx of people staying home. Calls to customer service centers skyrocketed along with demand for more intelligent ways to predict customer and infrastructure needs.

The situation urgently called for AI. Growing beyond just a shorthand way for telcos to triage customer service, over the past year AI has expanded into a way to coach customer service agents through more personalized interactions, to collect cell phone tower data to predict failures hours in advance, and more.

A recent IBM Institute for Business Value (IBV) global study found that 97% of communications service providers (CSPs) using virtual agent technology for customer care reported a positive impact on customer satisfaction. Another recent IBV study of telco CEOs found 58% view customer insights as most responsible for driving their success over the past three years. Our own usage trends underscore this shift as well: IBM Watson Assistant had 60% growth in monthly active users from January to December 2020.

Put simply, AI has become a critical, central tool for telcos to deliver deeper, more meaningful insights that can enable more informed decision-making and predictive action, from customer experience to infrastructure improvements.

Expanding real-time, predictive decisions

2020 was a stress test on network bandwidths. Accelerated by the pandemic, telcos are increasingly using AI to put more intelligent solutions in place to help turn insights into action and value.

For example, multinational telco Telefónica created an AI-infused platform that enabled predictive insights from cell towers in Argentina, enabling the creation of data models that can predict outages and failures before they happen. Telefónica is now able to predict failures within 24 hours, better allocate maintenance spend and reduce their response time to failures by a few hours — all leading to improved customer service at a time when reliability is of critical importance.

AI-powered, analytics-driven predictive modeling is a key way that businesses are expanding how they improve customer experiences by shifting focus from reaction to proactive action.

Trading in hold music for more insightful customer service

Over the last year, many consumers stopped visiting retail stores for help with changing service plans or gaining additional insight. The shift away from in-person customer service led to an urgent need for more complex digital virtual assistance.

AI-powered virtual agents went from more basic, shorthand triage to being able to handle more complex customer questions about topics like billing, usage upgrades and service changes. Additionally, telcos began to use AI to help customer service agents gain more informed insights on the customers themselves, so interactions could become more personalized and efficient when a customer opted to speak with a live agent.

Recently TIM Brasil, the Brazilian subsidiary of Telecom Italia, worked with IBM to transform their customer service workflows. Together they implemented an AI-powered virtual assistant to help handle the large volume of customer calls while maintaining customer satisfaction, reducing costs and freeing up human call center agents for higher-value work. TIM’s virtual assistant uses IBM Watson Assistant on IBM Cloud to process natural language interactions, responding by voice in real time to consumers’ questions or problems on issues such as plan benefits and bill payments.

Within four months of augmenting human customer care agents with the virtual agent, TIM’s containment rate increased to 75% and first-call resolution increased by 84%.

Building seamless experiences across the enterprise

As CSPs continue their adoption of AI-powered virtual agents, the more deeply they can integrate the technology into their existing enterprise systems and workflows, the more benefits they will receive. According to the previously mentioned IBV study of telcos using virtual agents, 67% of CSP survey respondents have already achieved their return on investment, and 17% have exceeded it.

Many early adopters have achieved success by expanding their applications of AI internally and externally to drive more intelligent workflows across their organizations. In other words, in order to glean more insights from valuable — and often untapped — data, it is important for telcos to consider creating an enterprise-wide AI strategy that prioritizes vertical and horizontal implementations.

One way to do this is to add capability within a vertical domain like customer service — for example, adding more customer intents to increase the kinds of customer queries that a virtual agent can handle by itself. Another option is to look horizontally across other internal processes and make those workflows more intelligent with the help of AI. For instance, a CSP that started with implementing AI-driven virtual agents in customer service might choose to apply the technology to human resources workflows, helping to answer basic employee questions about payroll, benefits and company policies and freeing up HR leaders’ time to perform higher-value talent strategy work.

While 2020 began with so many unknowns for businesses, consumers and IT leaders, it became a year of unprecedented digital transformation and intelligent insights. AI was critical for telcos in powering this transformation, and it has matured into an essential tool for enhancing business and customer outcomes in the years to come.