How AI Is Revolutionizing Cost Efficiency and Customer Experience in Telecom

The History of AI in Telecom

AI has been utilized in the telecom industry since the early 2010s, primarily in helping communication service providers (CSPs) reduce costs. AI began in telecom with predictive solutions that leveraged structured data. Common use cases have included anomaly detection (from a cybersecurity, fraud and network performance perspective) as well as chatbots for basic customer care tasks.

 

Employing predictive AI to optimize energy consumption has become a more common use case following geopolitical events and the impact of the COVID-19 pandemic, particularly as CSPs have become more focused on reducing costs. CSPs are also under pressure in certain markets to align with environmental, social and governance (ESG) agendas, and AI has emerged as a technology that can help CSPs reduce their carbon footprint.

 

The next stage of AI evolution pertains to generative AI (GenAI), which leverages unstructured data and opens up a broader range of use cases for CSPs.
 

Join TBR for insights into the current state of GenAI adoption in the telecom industry and the types of business outcomes the industry can expect from AI and GenAI — Save your spot today

GenAI Promises Better Customer Outcomes and Cost Reduction of Up to 80%

TBR’s latest telecom research indicates customer care, which includes contact centers, will be profoundly transformed by AI. Though traditional AI has been utilized in customer care for many years (e.g., chatbots and interactive voice response), GenAI will take customer care to an advanced state.

 

TBR estimates that GenAI could reduce the costs of running contact centers by up to 80%, and this is an area telecom operators are keenly interested in as they remain focused on cutting expenses across their businesses.

 

GenAI can also optimize the customer experience, essentially creating a super agent that is able to handle more complex tasks and lead the customer to better outcomes, thereby reducing churn and potentially driving upselling and cross-selling opportunities.

CSPs Will Expect Vendors to Share Cost Savings Realized From the Use of AI

When vendors are able to bring true AI capabilities and solutions to CSPs, the CSPs will find value in a variety of use cases such as AI-based applications for network maintenance or optimization, which will increase the efficiency of network operations.

 

Vendors will also benefit from cost efficiencies gained from AI, but TBR’s research indicates that CSPs will expect vendors to pass along some of the cost savings from the use of AI, such as costs freed up from a reduction in human resources.

 

Automation, analytics, AI and machine learning technologies will all prove critical to helping vendors improve margins during the 5G era and beyond. Examples include portions of Nokia’s AVA (Analytics, Virtualization and Automation) portfolio and Ericsson’s Operations Engine.

Most CSPs Will Remain Technology Consumers, Not Technology Producers of AI, Limiting Their Ability to Generate New Revenue in This Area

The 2 Primary Ways CSPs Will Derive Revenue-related Outcomes from the Use of AI

Leveraging GenAI to cross-sell and upsell existing subscribers may provide optimal revenue capture on a per-subscriber basis, while on the revenue protection side, CSPs will likely use GenAI to improve churn by better addressing customer pain points and root causes leading to the decision to switch operators.

 

The primary location where GenAI technology will be incorporated is the CSP’s contact center and potentially during the digital sales journey, such as interaction with a GenAI-enabled chatbot in the CSP’s digital storefront.

Most CSPs Are Expected to Rely on the Vendor Community and Hyperscalers for AI Innovations

CSPs will rely on vendors (including hyperscalers) to provide, and in many cases support and manage, AI solutions in their operations. For example, traditional network solution providers like Nokia and Ericsson as well as disruptive network solution providers would bring AI innovations as it pertains to the RAN and core to CSPs, in many cases incorporating AI innovations into software and processes.

 

Meanwhile, hyperscalers would likely be the de facto foundational large language model and other types of AI model providers, on which CSPs would leverage for their telco-specific use cases, such as in the areas of network operations, contact center and customer journey.