Telecommunications and other technology companies are under significant cost pressure as revenue has stagnated, with total telco market size outpacing inflation in only 2 out of 29 European countries in the last 15 years. To maintain competitiveness and continue to deliver value, telcos need to pursue a radical efficiency agenda of next-generation levers unlocked by emerging technology.
Many modern telcos are engaged in this already, deploying AI-driven solutions across the operating model, including customer support and network management. Done properly, such solutions dramatically cut response times, increase precision, improve customer and employee experience, and reduce or eliminate redundancy and waste driven by legacy systems and processes. More than just technology, this transformation is a complete reimagination of service delivery. Scaling these solutions across risk and reward gives leaders serious opportunity in areas such as customer intimacy, prioritization of where and how to change, and also in buy-or-build considerations.
When telecommunications industry leaders gathered at a recent event, the discussion they shared indicated a clear ambition to scale AI and generative AI in service operations across both telco and tech operations. At the same time, however, they are concerned about the corresponding change management needed within an organization, the quality of data that informs the AI systems to be used, and the impact of balancing cost with ROI. With the potential to transform customer engagement, improve service and field operations with personalized communication, and procurement processes with cost optimization, how can leaders use generative and agentic AI tools as a catalyst for rewiring their service operations to become both more efficient and customer centric?
From these discussions, three areas of decision-making emerged for leaders tasked with deciding what and how to scale: What is your level of risk appetite? Where to start and how to change? Is it better to build or buy new tools?
Risk and reward
One of the most significant advantages of integrating generative and other forms of AI into service operations is the marked improvement it can bring to customer experience and response times. By analyzing extensive data sets, gen AI can anticipate customer needs and tailor personalized solutions, fostering a proactive approach that translates to heightened customer satisfaction and loyalty.
With risk appetite, there are three trade-offs around which decisions can be made. The first centers on customer intimacy and which interactions should be AI-led versus human-led, depending on the customer’s mood and the importance of the transaction. For example, handling a customer complaint might be best resolved with human interaction, but a gen AI bot might be faster than a human at providing transactional support.
When it comes to IT, data investment, and the pace of rollout, the decision points center on whether to take on incremental or transformational change, and whether to pilot first or undertake full-scale deployment. At the event, discussions in this area focused on the challenges of unstructured legacy data, the need for courage to build systems anew, and the preservation of data privacy and security as it directly influences customer trust.
Technology risk can be reduced with investment in responsible AI, ensuring both ethical and legal compliance as well as reducing risk exposure due to unexpected model behavior or hallucinations. The preservation of data privacy and security is paramount, as it directly influences customer trust.
Where to start, and how to change
Rewiring service operations with generative AI requires a collaborative effort. Success depends on fostering a culture of innovation and continuous learning, which involves not only adopting new technologies but also transforming organizational structures and processes. Leaders need to think about two main areas of decision-making here. The first is domain prioritization (a subset of your enterprise that encapsulates a cohesive set of related activities), and the order in which domain transformations are undertaken and scaled to capture value while minimizing both risk and workflow disruptions. The second is simplification, and whether legacy processes should be automated as they are, or simplified prior to automation.
To make or to buy?
Decision points around differentiation, talent, and vendor concentration are all necessary when it comes to making decisions around whether to build or buy the AI tools needed. AI-powered chatbots and virtual assistants are now capable of crafting messages that resonate with individual customers. By leveraging natural language processing, these tools can comprehend customer queries and deliver pertinent, personalized responses. This level of customization has the potential to greatly enhance customer experience and lower service costs, and is a key differentiator that may only be available by building or customizing a solution, rather than deploying off-the-shelf tooling.
Make-versus-buy decisions also apply to talent: Should organizations hire external talent in a competitive market, or instead focus on building the skills of their own teams? As with technology, this is not a simple either-or decision, as most organizations will need to deploy a mix of both external talent and upskilling of existing talent, fostering a culture of change and AI adoption to unleash the full potential.
With a proliferation of technology vendors, and the need to integrate with legacy systems, the final decision point is whether organizations should focus on building technology with preferred vendors for speed, or distributing to reduce switching costs in a rapidly evolving landscape.