By: Albrecht von der Recke

The sheer amount of data that we produce and consume is exploding at an unprecedented rate globally. Data growth analysis shows that most of the world’s data today has been generated in just the past two years. On average, every human created at least 1.7 MB of data per second during 2020. By 2025, 463 exabytes of data is expected to be generated each day by the global population.

The pace is only going to accelerate. Ubiquitous smartphone adoption and ever-increasing consumer dependence on mobile devices are the major growth drivers, with mobile data volumes predicted to increase tenfold over the next five years.

With mobile users creating billions of touchpoints each day, highly valuable—and in many cases actionable—data points are flowing across every operator network. Yet, according to Forbes, less than 0.5 percent of that data traffic is properly analyzed to extract any additional value. A key reason for this is the unwieldy weight of the traffic concerned. A typical network in South America, for example, will have some two billion data events occurring every single day. Until now, cost-effectively analyzing these data points to create customer insights has proved to be an insurmountable technical and operational challenge.

Enter AI and ML: cost-effective data insights

In the past, operators have primarily marketed to their customers along traditional campaign lines using mass, non-personalized SMS offers to purchase top-up or airtime bundles, perhaps sent once a month. Typically, conversion rates for such campaigns are below one percent. But recent advances in data visualization are offering the opportunity to completely transform operators’ understanding of customer behavior and buying patterns. The huge volume of data that users are generating on their mobile devices today, combined with new AI and machine learning-led capabilities in data analytics, means that operators can now gain more insight than ever before into what customers are doing, what their needs are, and when is the optimal moment to get in touch with them. This new ability to microsegment and micro-market to customers based on real-time requirements—whether for top-up, small loans, or mobile wallet transactions—can deliver conversion rates of up to 10 percent, meaning a more than tenfold increase in marketing success for operators and a significantly improved user experience for mobile customers.

The great technology leap forward in operators’ ability to visualize these billions of daily customer touchpoints has been cloud-based machine learning and AI, which provide the ability to manage, process and analyze data in much more economical ways than previously possible. Machine learning and AI offer the processing capability to find highly targeted “needle in a haystack” data at cost levels that make business sense for operators. The key to unlocking the potential of data has always been whether it can be processed at a cost point that allows a profit, and AI and ML are finally delivering the elasticity that is making operators’ big data ambitions a reality. For example, to examine a base of 50 million customers, the cost of analyzing all the customer datasets generated and producing actionable results may now be around only €1,000-2,000 per day which, from an operator perspective, represents an attractive business model.

Know your customer: right offer, perfect time

What datasets can be analyzed and how can these be monetized to boost revenue for operators and better meet customer needs? It all comes down to granularity and real-time analysis into what the customer is doing in that moment.  All networks offer a range of static datapoints, such as what handset a customer is using, their typical data spend and other standard CRM insights. But the ability to add real-time data about a customer’s interaction with their device creates the opportunity to detect patterns and build segmentation that can form the basis of much more effective marketing campaigns, and create happier, more loyal customers.

Every time a phone tries to connect to the operator network—to check an email, load a TikTok video, or look at sports results—a data event is generated. Operators can look for general patterns in these events to deliver timely offers that could be more tailored to potential interest areas, such as football, gaming, business and so on. As an example, when the football World Cup tournament is on and data usage might be higher, operators could offer prepaid customers who are running very low on data credit a short-term data pack that allows the customer to stay connected. With five billion prepaid customers around the world—and each engaged in an ongoing natural lifecycle of running out of data or airtime—the ability for operators to instantly offer top-ups, product bundles or even extend small loans to ensure continued connectivity provides huge value to both customer and service provider.