Business Intelligence and Analytics –The next frontier for African telecom operators


Opportunities and rapidly changing market dynamics make Africa one of the most vibrant markets in telecom.  Grand entry of MNC’s like Airtel, Vodafone and innovative initiatives like M- Pesa by incumbent players have brought the African telecom market to the limelight.

Opportunities and rapidly changing market dynamics make Africa one of the most vibrant markets in telecom (image: stock.xchng)

Africa, with 735 million subscribers, is poised to reach 911 million subscribers by 2014. The verdict is out and clear, communication in Africa has become a basic social need across all stratums.

There are early evidences that the telecom subscriber base in Africa will have an urban versus rural divide. The behaviour and usage patterns are likely to be different in these segments. While in rural markets, the key factor is enabling basic connectivity; in urban markets it is the drive to have infotainment and convenience by a more tech savvy population.

Hence the usage of value added services (VAS) would be on the rise. Such services include traditional VAS services and data heavy services. Some of these newer services are Mobile games, location based applications, movies on demand etc.   The impending launch of Mobile Number Portability in many countries would fuel the migration of subscribers to better and cheaper alternate operators. This will force the operators to understand the needs of their higher value customer segments better for retention and increased loyalty.

Technically, the projected growth in telecom generates huge amount of data for every transaction that happens with a subscriber, as well as data generated through other processes of the telecoms. But the striking truth is more data necessarily does not mean more intelligence or better decisions. More efficient way of processing of large data and deriving effective information / knowledge for decision making is the ultimate function of any BI/BA effort in a Telco. In view of the above trends of urban-rural divide and MNP, the need for effective BI and analytics would be paramount in the coming months.

What do African telecoms need today?

In a Telco setup, deeper understanding of the subscriber base and predicting their behavioural characteristics well in advance will keep the company on a profitable path. In turbulent markets like Nigeria, Tanzania & South Africa where multiple operators are competing with each other, this requirement becomes much more important as the adage ‘whoever controls the future will stay in businesses’ aptly suits the scenario.

Telco has the classic 80/20 distribution when it comes to its subscriber base and revenue as shown in the long tail chart. Since the higher revenue comes from a small percentage of subscriber bases, any loss in that group will result in drastic drop in revenue. On the other hand when the customers in the long tail are not properly handled, it may lead to drain on revenue. Hence telecommunications need a unified decision support system, which can provide the analytics at different levels as per the need. This gives rise to the need for a next gen BI/BA system explained in the following sections.

Why go for next-gen BI tools?

The last generation of Business Intelligence and Analytics were technology focused. Each component of the BI architecture/stack was sourced from multiple vendors resulting in an integration nightmare.  Recent consolidation of the BI market by the majors like IBM, Oracle, SAP and Microsoft has given a fillip to a single vendor stack for all software components and a clear roadmap of future developments. This augmented by more demanding and involved business users as well as increasing competition among telecoms has created a need for unified approach to business intelligence and analytics in telecom space.

The next generation BI/BA for telecom will be different from last generation systems both in terms of technology as well as the functional features. Gen X BI / Analytics was limited to privileged few while the next generation – Gen Y BI/Analytics will be pervasive, enabling the information democracy by providing relevant decision models to arrive at a better decision using insight information at hand insights to everyone at their level.

What are the applications of unified decision support systems in telecom?

The applications of the unified decision support system span across all critical operational areas of telecom like

* Marketing, CRM & Sales

* Fraud Detection

* Network Management

Marketing, CRM & Sales

The  large amount  amounts  of  customer  transaction  and  call  detail  data,  collected  in the  operational  systems  of telecommunications,  provide  valuable insights to understand and predict consumer behaviour.

* Profiling customers using CDR

* Measuring customer value and retaining profitable customers

* Maximizing the profit obtained from each customer – Cross sell, Up Sell

* Targeted Customer Acquisition

* Observing patterns in consumer and industry trends

* Churn prediction: Predicting customers’ propensity to churn

* Churn management:  Identify underlying causes for churn and take appropriate actions

Fraud Detection

Fraud is a serious concern for telecom companies globally leading to a revenue loss for the operators and the burden is being passed on to the existing customers. Using predictive modelling and analytics fraud can be controlled. There are a lot of statistical techniques like   rule based analytics, clustering, Bayesian rules, visualization methods, and neural network classification equipped to handle fraud adeptly.

* Identification of potentially fraudulent users and their atypical usage patterns (subscription fraud)

* Detecting attempts to gain fraudulent entry to customer accounts (superimposed fraud)

* Discovering unusual patterns that may need special attention such as busy- hour, frustrated call attempts, switch and route congestion patterns, etc.

Network Management

Growing and maintaining profit margins in the Telecommunications industry requires optimum network efficiency and ensuring high network reliability. BI and Data Mining analytical tools have proven to be very effective for comparing a wide range of metrics across network operations, creating real-time reports for  identifying problems  that need  immediate attention, and generating alerts for  instant  notification  of  emergency  situations  requiring  rapid  response.

* Network fault identification

* Alarm correlation (for relating multiple alarms to a single fault)

* Network fault prediction

* Identifying and comparing data traffic

* System work load management

* Resource usage management

* User group behaviour


The Next gen BI will see an increased role in the telecommunication compared to the last generation BI and this is the consequence of the competitive landscape and business user focused BI. While the product and services differentiation are becoming blurred, companies need to differentiate themselves from the rest by ore effective use of the transactional data created.

This will give telecoms the much needed competitive edge by helping them to cut down operational costs, reduce revenue leakages, understand consumer behaviour and offer better products & services. Hence BI & Analytics are increasingly vital today for a telecoms sustainable growth in a dynamic environment.

Dr. Jay B.Simha, Chief Technology Officer of ABIBA Systems

Comments are closed.