Globe Telecom: Real-world AI in customer experience transformation

Globe Telecom is the largest mobile operator in the Philippines, with reported revenue for 2023 of P162.33bn ($2.89bn) and 54.7 million subscribers as of November 2023. Most of these mobile users are prepaid subscribers, which makes for a highly competitive market where users can — and do — switch carriers frequently. The barrier for switching is very low, and customers often jump to take advantage of short-term (typically one, three or seven days) promos.

Globe needed a way to stabilize and improve customer retention. To do so, the company first needed to "dimensionalize" its customers. Current metrics, such as the Net Promoter Score (NPS), have limited use as a standalone number. However, when correlated with internal data, NPS becomes more valuable and can be used, for example, in modeling for churn prediction. It becomes especially useful when combined with internal data and qualitative feedback and powered by natural language processing (NLP) and artificial intelligence (AI).

Like all carriers, Globe has a rich amount of data, but it needed help to readily leverage this data on a per-subscriber basis for troubleshooting subscriber issues, devising new promotions, creating new bundles, etc. Globe is interested in taking these capabilities one step further than its competition by analyzing and leveraging data quickly and with more granularity. The timing of the data is critical, as Wil Sarmiento, CRM head at Globe Telecom, remarked to me in an interview, "If something bad is happening, you don't want to respond next week or even the next day." To turn around the situation and attempt to retain the customer, the communications service provider (CSP) must respond quickly — ideally with a combination resolution/offer that is personalized and meaningful to that specific customer.

The AI journey

Globe must harness both structured and unstructured data from multiple channels to create a personalized image of the subscriber, and it must do so without a time lag if the company hopes to apply its learnings to customer experience (CX) on the fly. AI is critical to helping Globe meet these goals. To create an AI-powered CX proof of concept (PoC), Globe has turned to two partners, Amazon Web Services (AWS) and Stellar Elements, an Amdocs company.

Globe tapped AWS for its Connected Customer Journey (CCJ) platform, part of the AWS Solutions for Telecommunications product portfolio. CCJ, as its name suggests, enables CSPs to centralize and refine customer-context data from multiple channels. By taking siloed, raw and incongruent data and then analyzing it in real time, CCJ can deliver to the CSP a "single source of truth." The CSP can then use that to create an image of the customer journey and derive suggestions for improving that journey.

The AI-enabled CX PoC is structured around a robust data strategy, with Stellar Elements building data products on a data lake built on AWS. None of the large language models (with prompts) leave the Globe account, and the base model is trained only on Globe data. All of this adds to both the security of the solution and the ability for hyper-personalization. The resulting solution can provide capabilities such as the following:

• Real-time churn prediction

• Real-time customer journey segmentation

• Next-best experience prediction

• Personalized communications and marketing

• Human-like chatbots

According to Ajay Ravindranathan, principal AI/ML solutions architect at AWS, "Utilizing capabilities like Amazon Neptune for easily modeling complex data relationships supported by Amazon Neptune ML and Amazon SageMaker integration enables customers to focus on their data and extract the most value out of it."

Next steps in the journey

Globe already places significant emphasis on CX, but the new AWS CCJ and Stellar Elements solution will allow it to "get to the next level," pushing the envelope and identifying new opportunities. The journey is not without its challenges, however. Globe identified one of its most significant issues as the existing organization, both people and systems. As Sarmiento said, "Change management is hard. The day-to-day jobs change as we re-tool the Globe CX system." All elements of the organization are involved as Globe works to bring in data from multiple sources, including multiple operations and business support systems and call centers. Bringing all this together at scale while maintaining security and reliability is difficult.

Looking ahead, Globe is building toward a closer collaboration between the care side of the house and marketing. The care side of the house creates a lot of valuable data that can be used to deepen relationships with individual customers and engage them in meaningful ways across the lifecycle. Globe identifies the use of this data as a "huge opportunity," as it generates incremental value for the carrier and its customers. As its CX journey unfolds, Globe is eager to leverage its ability to reveal the connected customer journey of its customers, individually, across channels and over time.

This blog is sponsored by AWS.