Unleashing the Potential of Streaming Analytics for CSPs

Last Wednesday I presented a webinar in conjunction
with TM Forum on the benefits associated with streaming analytics for CSPs. During the presentation we covered a variety of topics including the important differences between batch and streaming analytics, how streaming analytics fits into the big data ecosystem, what problems and opportunities can be addressed by streaming analytics, and real-world use cases illustrating the power of streaming analytics in action.

Here are a few of the highlights and key concepts featured in the presentation:

The Difference between On-Demand Analytics and Streaming Analytics

Streaming analytics provides continuous analysis of events as they unfold. By contrast, on-demand solutions are static and batch-oriented, meaning that the data flows into repositories for “snapshot” analysis. The fundamental difference on a practical level lies in the data latency associated with batch processing. If data is sitting in repositories waiting for on-demand analysis, issues are detected after-the-fact when it’s too late to prevent the ripple effect throughout the broader organization and the impact on the customer.

Complementary Big Data “In Motion” and “At Rest”

While batch processing solutions by definition are limited around their ability to perform continuous monitoring and analysis, Big Data frameworks such as Hadoop provides impressive data storage capabilities and the ability to mine prodigious quantities of data for deep insights. Vitria leverages Lambda Architecture, incorporating Hadoop as a batch layer for historical analysis and long-term data storage. The speed layer – powered by streaming analytics – provides continuous analysis and automated remediation within seconds of occurrence. It’s the combination of the benefits of Hadoop-based storage and historical analysis capabilities and streaming data analytics that makes Vitria OI so powerful in Lambda-style architectures.

The Real-World Problems and Opportunities that Streaming Analytics Addresses for CSPs

There are a wide range of benefits and capabilities that streaming analytics delivers for CSPs, including real-time:

  • Network Optimization – continuous monitoring of performance across all network cells for issues such as dropped calls with drill-down to individual groups of cells, individual cells and individual issues; adjacent cell performance, unreported data detection, and predictive failure analysis
  • Customer Experience Management – continuous monitoring of individual customer experience including detection of service-affecting events such as dropped calls for VIPs and corporate and roaming VIP customers with “track and trace” capabilities
  • Fraud Detection and Security – mobile originated spam detection and mobile wallet fraud detection
  • 1:1 Marketing – personalized, relevant and timely marketing offers delivered to customers based upon where they are, where they have been, and predictive capabilities on where they’re going next
  • Revenue Optimization – analytics to support new account offerings, dynamic top-ups and use-based charges and mobile data pricing
  • Internet-of-Things and M2M Initiatives – as operators seek to exploit their networks to monetize opportunities in smart metering, asset management – eHealth and connected car projects, for example – operational analytics becomes an essential tool

While it’s important to understand that on-demand analytics continues to deliver significant benefits, the key takeaway from the webinar is the vast range of capabilities that streaming analytics delivers to CSPs in a manner that is entirely complementary to on-demand. The possibilities are endless.

One thought on “Unleashing the Potential of Streaming Analytics for CSPs

  1. Effective use of Mobile Analytics has the potential to open up a unique set of business opportunities in the Indian Telecom Industry. Knowing how customers interact with your mobile channel is vital to the success of a mobile strategy.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>