YOU ARE AT:OpinionReader Forum: Better risk management through big data, Hadoop

Reader Forum: Better risk management through big data, Hadoop

Editor’s Note: In an attempt to broaden our interaction with our readers we have created this Reader Forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].

Every business wants to stop bad things before they happen. Risk management doesn’t provide us with that superpower. What it can do is gauge the odds that a particular event will happen, monitor the effects of its occurrence, and – at its absolute, most basic level – help businesses understand if the risk presented is acceptable, can be mitigated, or should be avoided at all costs.

Those in the telecommunications business have long been among the most avid collectors of data. The problem for communication service providers was never having enough information to analyze; it was finding tools that could deal with vast volumes of data in a useful way. Data storage costs were quickly eating up budget allocations as CSPs struggled to do more than the most rudimentary analysis.

For a while, it looked like big data would become another much-hyped project that failed to live up to its predicted potential. But the techiest among us knew there was business gold to be mined from the data heaps. So Google engineers developed a way to store and process vast, swift-moving streams of data. After releasing a white paper about their innovations, their solution turned into an open-source project called Apache Hadoop. Hadoop removes many of the barriers that businesses had encountered in collecting and utilizing big data effectively.

Apache Hadoop and the telecom industry

Hadoop is both a data-storage and analysis platform. It provides large-scale storage of any type of data in a cost-effective way. Additionally, Hadoop’s utilization of commodity hardware clusters provides the optimum environment for long-term growth and return on investment.

Hadoop also enables highly granular analysis of unstructured and structured data. As such, Hadoop is of particular benefit to telecommunications companies that typically store vast quantities of data collected from customers and network infrastructure sensors.

Previous data analysis solutions enabled sampling of vast telecom data lakes, rather than the deep-dive abilities provided by Hadoop. Being able to analyze data stores for historical and current patterns is critical to risk-management processes, such as fraud detection, customer churn and network performance. And Hadoop stores data in its native format, so information that may be of use tomorrow or next year does not get lost in translation.

Hadoop, big data and risk management

Sophisticated risk management relies on context. While one could make a prediction on how profitable and loyal  customers are likely to be based on their accounts payable records, an analysis that includes products and services viewed, reviews and comments on social media, customer support requests and preferences, as well as bill payment histories reveals a much richer and more actionable picture.

The analysis becomes even more valuable when customer groups are created based on preferences, and accurate predictions can be made based on the profile of a representative section of the group. This enables a CSP to proactively reach out to valued customers at the right time and in the right way – far more effective than reinventing the retention wheel for every customer.

That said, customers act in mysterious ways. Hadoop’s ability to provide speedy analysis of the freshest data can reveal sudden disruptions – perhaps in response to a new service offering or recently launched product. Looking at newly created data can be a significant game changer that dramatically turns around a potentially difficult launch or initiative. Why wait until a crisis occurs to conduct crisis management? By the time a brewing issue begins to manifest in social media posts, it may be too late to rescue.

Deep understanding of customer patterns can also help protect CSPs from loss. As the industry moves toward what is essentially a leasing model for devices, it becomes important to base upgrade offers on factors that extend beyond a customer’s recent payment history. The same holds true for service packages. A company may find that the exposure they gain from enabling early adopters’ access to devices and services outweighs the fact that some of their most eager customers may have less than sterling credit scores. Hadoop enables CSPs to gain a more accurate view of danger and gain, so that they can manage risk more intelligently and improve revenue through thoughtful customer retention actions.

Customer retention extends to understanding whether your customers are still with you. Switching service plans to get more value for their money has become a common customer activity. This can complicate a CSP’s understanding of whether customers are leaving because they prefer a competitor’s prices, network coverage and services, or are happy customers who simply wanted to change their service plan.

Keeping the network ready, steady and safe

Sensor data is an invaluable tool in keeping the network up and running at optimum performance levels. Historical data and maintenance information can be used to predict hardware failures and system overloads, as well as to develop an understanding of how a failure will affect the capacities and health of the network as a whole.

Sensor data can enable long-term trends and evolving problems to be spotted, addressed, or at least mitigated before their effects became apparent to end users.

For large-scale disaster planning, providers of essential services can look at how their systems interact and rely on each other in order to put the necessary redundant and emergency systems in place. Real-time monitoring can be conducted during crisis situations to help maintain critical services. Forensic analysis of large data volumes can be performed after the fact to find victories to build on as well as ways to improve.

The world of CSPs, utility companies and first-responder teams are becoming increasingly integrated and dependent on each other. Using a solution that can perform rich granular analysis on all data formats, at a huge scale, will improve efficiencies, mitigate and manage mission critical risks, and – most importantly – save lives.

Sameer Nori is senior product marketing manager at MapR Technologies. Nori has 10-plus years of experience in the technology industry in marketing, sales and consulting. With an executive MBA from the Fuqua School of Business, Duke University, Nori’s domain of expertise is in business intelligence and analytics.

ABOUT AUTHOR

Reader Forum
Reader Forumhttps://www.rcrwireless.com
Submit Reader Forum articles to [email protected]. Articles submitted to RCR Wireless News become property of RCR Wireless News and will be subject to editorial review and copy edit. Posting of submitted Reader Forum articles shall be at RCR Wireless News sole discretion.