An expert guide to the current state of the telecom industry and data analytics’ role in its future.


The telecom industry is complex and exciting. It’s dynamically growing as companies move into related services. And it’s profitable, projected to hit $ 1.6 trillion USD in 2020 [Source: IDC]. 5G network infrastructure revenues worldwide will touch $4.2 billion in 2020, as predicted by Gartner [Source: Deloitte].

But telecoms are also data powerhouses. A typical mobile provider (~80 million subscribers) generates close to 2.5 quintillion bytes of data every day, and the volume of data is set to double every 1.2 years.

With all this data and all the moving parts involved, telecoms are also very complex. That’s both a challenge to deploying analytics and a very good reason to do it!

In this article, we’ll look at the current telecom landscape – its players and key developments – and examine how data analytics can successfully be applied. We’ll also consider some emerging trends to watch.


The Telecom Landscape

Telecom is much more than voice calls. It includes all kinds of information – signals, messages, texts, images, sounds – transmitted by wire, radio, optical, or other electromagnetic systems. We usually think of telecom companies primarily as service providers, but the market is actually far more complex. It comprises:

  • Mobile Network Operators (MNOs) like Verizon, Sprint, Jio, etc.
  • Mobile Virtual Network Operators (MVNOs) like Boost, Virgin, XOX, etc.
  • Tower Site Owners/Operators: MNOs and MVNOs often lease tower sites from other companies who own, manage, and maintain them. For example, AT&T has a service agreement with American Tower, an independent owner, operator, and developer of communications real estate; India’s Indus Towers is owned by a group of communications companies.
  • Equipment Manufacturers, like Nokia, Cisco, and Huawei, that build telecom equipment, smartphones, and other consumer devices.
  • Cable Operators like Charter and Dish Network.
  • Over the Top (OTT) services like Netflix, WhatsApp, and Spotify that are delivered outside of cable, satellite, or broadcast TV.
  • Mobile Money (MoMo) services like PayPal and Alipay that allow people to spend and transfer money via mobile device.

In this market, each player has its own analytics requirements and nuances, and bringing them together is no easy task. Fortunately, we have data analytics on our side.

There’s also an interesting divide between telecom leaders in different areas. Companies in established markets (e.g. Western Europe, North America, etc.) tend to be ranked by value, while companies in emerging markets are ranked by subscriber count. Top value companies focus on retaining customers, cross-sell and upsell opportunities, and cost optimization; top subscriber companies are more focused on growth.


Key Developments in Telecom

The last few years have ushered many changes into the telecom industry. For the first time, mobile is beginning to overtake TV as the biggest source of ad revenue [Source: Livemint]. There’s also a lot of interest around advanced analytics, 5G, artificial intelligence, and other modern technologies. We’ll only focus on two areas here, and you’re probably already familiar with them:

5G – the next big thing

Plans to deploy 5G networks were interrupted by COVID. But, prior to the pandemic, this was the hottest topic in telecom. Why so much interest? 5G is faster, but the real change is for businesses and the indirect consumption of data (e.g. smart cars, connected utilities, IoT, self-driving cars and machines, etc.). It’s also more reliable, with about 1 millisecond of lag time. This greater efficiency was projected to start an economic boom, as it makes more kinds of telework feasible.

The COVID effect

Again, this is not exactly headline news. COVID might have spurred tremendous growth in services like Zoom, but it’s also disrupted the telecom industry like nothing else. Network, voice, and data traffic have skyrocketed, forcing companies to re-plan network traffic. Customer service calls are up, but new subscribers are down. The industry has had to quickly adapt – providing extra data allowances, delaying phone launches, tying up with telehealth, e-learning, gaming, and OTT services, and deferring 5G rollout plans.

Additionally, some countries have mandated that telecoms share data with their respective governments; others have lifted restrictions on some types of communications, like VoIP.


Analytics Use Cases in Telecom

Like other industries, telecoms seek operational excellence, revenue maximization, and superb customer experience. And, also like other industries, analytics can help them reach these goals. So, let’s look at three types of analytics and what they can do for telecoms.

1. Customer analytics

Customer analytics is a boon for marketing. It can help telecoms create highly personalized campaigns and offers, basing these on specific interests and actions. In fact, customer analytics and telecoms are an ideal pair, as communication service providers have an especially rich vein of data to use.

Suppose you want to implement customer analytics from scratch. What needs to happen?

Step 1: First, you need to establish a complete customer view, i.e. compiling a Datamart with multiple types of data. For telecoms, this could include demographic, campaign, call record, customer service, sales, tariff, marketing research and survey, geolocation, and digital (DPI) data. We’re talking a huge variety, with potentially thousands of variables.

This complete customer view is the foundation of all customer analytics efforts.

Step 2: Once you have a Datamart, you can build appropriate models and deploy an integrated AI engine to provide action recommendations. Machine Learning, a subset of AI, is often used here.

Step 3: Finally, you need to get information into the right hands. For this, a user-friendly dashboard presents insights (e.g. acquisition, win back, growth, churn, etc.) and projects the expected impact if you take a recommended action.

For one telecommunications leader, we used a similar approach to help them target customers who were ready to upgrade from a feature phone to a smartphone. By targeting just 30% of their customers, we identified 75% of upgraders – all thanks to data analytics.

2. Customer Experience scoring

Analytics can also help telecommunications companies tune into problem areas. In this case, the data helped us develop very precise categories of customer experience, which were then scored and merged with network data. The result? The company could quickly see sites that were bleeding customers (showing high churn and a low customer experience index). They could use this to take decisive action and carry out predictive maintenance.

3. Network analytics

Traditionally, network analytics was based solely on network performance, but this leaves out a very important factor: the customer’s experience. Now, telecoms are merging network experience and customer experience for a value-driven approach.

By factoring in call rates, voice quality, call drops, and customer satisfaction rates, we can get a much clearer view of network performance. Thus, network planning (how many towers are needed, where to put them, what technology to put on them) and optimization (analyzing and optimizing spectrum usage, network uptime, downtime, KPIs, etc.) can be more aligned with company goals.

For one telecom company, we used a variety of advanced data analytics techniques – forecasting payloads, using game theory to determine next steps based on competitor actions, identifying financial metrics, network modeling, financial modeling, decision-optimization algorithms, and scenario creation – to help them plan and forecast their network optimization and planning efforts. And, of course, it had to be simple enough for business users to run what-if scenarios without recourse to the IT or data departments.

The result was simple: a map showing what to install, where to install it, and how much is needed – a simple thing with a very complex backend!

There are many other analytics applications that work well for telecoms – credit scoring, market mix modeling, churn modeling, etc. Pricing analytics is also great, albeit complex; the multitude of bundles, plans, and phone prices make this more interesting and more challenging. Geolocation-based analytics can also be used both internally (e.g. segmenting based on micro-locations) and externally as part of a data monetization program (e.g. supplying aggregate geolocation data to a retailer searching for new branch locations).


Expanding Opportunities for Telecoms

There are many new opportunities opening up for telecoms, but we will only go into two: mobile money and OTT services.

Mobile money is already on many telecoms’ radar, and it’s slated for even bigger growth. In terms of usage, services like ApplePay and PayPal are taking the place of cash. The same kinds of analytics we’ve already discussed can be used to encourage people to use these services (or use them more).

Over the Top services can also benefit from segmented consumers, AI-powered recommendation engines, optimized and personalized offers, and cross-channel offer delivery. For one such provider we worked with, personalized push notifications resulted in a 20% increase in user visits. Personalized emails with content suggestions gave a 50% lift in conversion.

And there’s also what Jio calls new commerce – connecting customers to merchants digitally. This can be a very interesting new avenue for telecoms.

In summary, we can clearly see that a partnership between telecoms and data analytics will be very fruitful. If you’re interested in learning more, don’t hesitate to reach out to our experienced team.

Authored by: Richa Kapoor, Marketing Manager at  Absolutdata and Promita Majumdar, Marketing at Absolutdata



Related Absolutdata products and services: AI & Data SciencesNAVIK AI PlatformCustomer AnalyticsMarketing Analytics