Big Data Mine for revenue

As voice revenues begin to plateau, telcos are increasingly looking at big data analytics to provide better customer services. Voice & Data takes a look at how telecos are leveraging big data

‘ Data, data everywhere… and every bit to be mined’ – That ’s the rhetoric of every service provider today. Be it structured or unstructured data, telcos are gradually, understanding the prowess of data and the role of big data analytics in improving their revenue margins.

While few years back, discounts, cash back offers were the strategies employed by telcos to woo their customers, these days with voice revenues about to reach saturation point; companies are increasingly looking at offering personalized services with customer retention plans, among others.

According to analyst firm, Information Data Corp (IDC), over 2.7 zetabytes of data exist in the digital world today and 90% of all data has been created in the last two years. The telecom industry generates structured, semi-structured and unstructured data from Call Logs, SMS logs, N/w log, Web Logs and so on… Mining these logs and Call Detail Records (CDRs) is helping telecos capture and understand key customer attributes.

Telecom operator Tata Teleservices has been using big data solutions for addressing legal and regulatory requirements which are mandated for longer retention periods. Given the growing volume of data and time sensitivity attached to such requirements organizations usually look to procure additional storage

Telecom operator Tata Teleservices has been using big data solutions for addressing legal and regulatory requirements which are mandated for longer retention periods.

and servers incurring high expenditure.

As an alternative, big data solutions are deployed which are cost effective in providing timely information to legal and regulatory teams. “While there are lot of open-source tools/solutions available, to meet the requirement and given the sensitivity of the data, Tata Teleservices deployed Enterprise Hadoop Solution backed by necessary security and support,” says Srinivas Rallabandi, head, business intelligence, Tata Teleservices.

Another telecom service provider Russian firm MTS is integrating and analyzing

information from call detail records of its customers and other company data sources on an Enterprise Data Warehouse (EDW) platform, which is helping it to track customers who are likely to reduce their use of MTS.

Besides, by analyzing social network topology, MTS is now able to identify opinion leaders and influencers. MTS uses this information to support viral marketing campaigns to sell value-added services. To execute the target marketing campaigns, the company uses the SAS Marketing Automation solution which is integrated with the Teradata Active EDW and social network analysis has proven to be more accurate than a traditional predictive modeling approach.

However, Rajeev Batra, CIO, MTS, believes operators are in the discovery phase of big data. “While analytics is being done extensively by all of them but big data is not used in a major way but only in bits and pieces I think, in a year or two we would see a much more big data adoption in the system,” says Batra.

Key Focus Areas

Industry experts are of the view that the telecom industry wants solutions around customer-experience management, network analytics, churn reduction, cross sell and upsell.

Telcos can deploy big data solution to benefit business and customers like marketing, customer interactions/ experience management, network management leading to increase in profitability. An uptrend can be seen in the deployment and adoption of big data analytics for business operations, primarily in customer experience management where there is a need to integrate structured (traditional) and semi-/un-structured data points. According to enterprise solutions provider IBM, telcos are going for two main aspects of big data analytics for their business growth –customer analytics and transforming operation.

The company has invested more than $1 billion in advancing new telecom offerings including telco-related acquisitions. It has signed agreements with telcos such as Bharti, Vodafone and Idea. For

Amdocs Big Data Analytics proposition is a holistic telecom specific approach comprising Data Science As a Service, Actionable Analytics Applications and Data Management and Modernization. —Hadas Haran product marketing director, Amdocs Big Data and Analytics Coupled with big data offerings of optimized appliances, IBM Cognos Business Intelligence software combines operational and financial data from across the organization into a single source of information for reporting, analysis, dashboards and scorecards to help drive better business decisions. —Asheet Makhija country leader, information management IBM India/South Asia

Vodafone, IBM India manages IT services, which include business intelligence, billing, and financial systems. Besides, IBM has enabled Idea to accelerate time-tomarket of new services and empower business with analytics tools, thereby improving end-to-end customer management and provide new revenue streams.

“Coupled with big data offerings of optimized appliances (Pure Data for Analytics and Pure Data for Operational Analytics), IBM Cognos Business Intelligence software combines operational and financial data from across the organization into a single source of information for reporting, analysis, dashboards and scorecards to help drive better business decisions,” says Asheet Makhija, country leader - information management, IBM India/South Asia.

Besides, IBM SPSS software encapsulates advanced mathematical and statistical expertise to extract predictive knowledge that when deployed helps improve outcomes, he adds.

However, according to Sunil Jose, managing director, Teradata India: “Network is definitely an important focus area for CSPs, there are three broad areas where big data can be used in network analytics—efficient roll out of the network which is the focus for CTOs, increasing revenue – focus for sales and marketing teams and, improving overall customer experience and brand building.

Teoco director and India country manager Srinivas Bhogle talks about a product called service assurance that gauges the network of different operators and if there is any problem it would raise the alarm and provide solution for the same.

Though most service providers are not starting entirely from scratch, having

already developed data warehouses and related business intelligence (BI) solutions, most realize that big data analytics require a different infrastructure than what they have used historically for data warehousing and BI. Many organizations, therefore, plan to invest in new solution infrastructure to realize the promise of big data. Enterprise Survey Group (ESG) estimates that more than half of larger organizations will make such investments.

Based on ESG’s modelling of a mediumsized Hadoop-oriented big data project, the preconfigured Oracle big data appliance is 39% less costly than a “build” equivalent doit-yourself infrastructure. And using Oracle big data appliance will cut the project length by about one-third, the company claims.

Oracle Big Data Appliance, in conjunction with Oracle Exadata, and Oracle Exalytics, helps customers to acquire and organize diverse data types, and then analyze them alongside existing enterprise data to discover new insight.

IMImobile, which operates in over 60 countries with over 100 operators, is investing on a product called profile manager, which works in close conjunction with another product called campaign manager. Profile manager is a way of aggregating customer data from multiple sources at one place without making changes to existing systems. So, there is no replacement to the existing systems but it works along with what it already has.

“Besides, we have chunk prediction to predict the probability of churn of given set of customer and if the model says that there is a high risk of churning then some retention activity can be undertaken for the targeted group of people,” says Sudarshan Dharmapuri, vice-president, product management, IMImobile.

He explains that while in India the analytics tools are just being used for retaining customers and offering better solutions, in mature markets big data analytics are used in a bigger way like –in analyzing the health pattern and in containing epidemics.

Hadas Haran, product marketing director for Amdocs Big Data and Analytics, elaborates on his product. He says, “Amdocs Big Data Analytics proposition is a holistic telecom specific approach which includes Data Science As a Service. SPs struggle to extract value from their data especially when large data sets are involved. Combining domain expertise around network, marketing, care and operations with data science, Amdocs DSaaS provides a valuable kick start to service providers by performing consultancy data discovery and situation specific analytical application development.”

Challenges

Experts believe quality is the biggest challenge as data is not at one place and the same data tells you different stories, so rationalizing data is one of the key concerns. According to Satyakam Mohant, director and CEO, Ma Foi Analytics, “The challenge lies in ‘variety’ not volume or velocity. If big data had been only about the amount of data being generated or the pace at which it is being generated, one just needed to scale up existing data storage and processing systems to take care of it.”

The main challenge is in integrating the relevant data which exists in the telco ecosystem but is distributed across teams, locations and formats says VVR Kishore, head – India operations at Mobileum, which has over 100 analytics engagements currently in India and across the world.

Big Data deployments or rather engagements typically start with “proofof-concepts”. The vendor and telco brainstorm on current business challenges and finalize a few use cases or examples of how certain types of data can be analyzed to deliver richer insights. The vendor then works with the operator to build the use case out and show interim results. If these are found to be compelling, the telco commissions a full-fledged deployment where there is a big data platform installed and integrated with multiple OSS/BSS data sources and the vendor begins work on implementing a set of use cases for the telco.

“Some of the common barriers to big data analytics are the same as one would expect with any new transformational concept. First, there are organizational barriers such as – Who owns it? Is it IT/ CIO, business/marketing, or a brand new ‘department’ for big data analytics? How do we ramp up the skills to design, establish and run functional big data analytics? Do we need to engage expensive consultants?,” points out Rallabandi.

Then, there are the inevitable RoI worries – sponsorship, business case, TCO. There is also the worry that the current data base is too messy, or unstructured, to lend itself to big data analytics. The fact is, this is not a limitation for big data analytics at all, he adds. One just needs to get started.