Improving Marketing Agility through Real-time Decisions
By : Nirav Bhatia
Director - Presales
Mobile subscribers expect value-driven engagement at every touch point. Communication Service Providers (CSPs) need to proactively manage subscriber expectations and offer personalized services for a seamless experience. This is where the marketing agility of a CSP is being questioned and the value of Big Data Analytics is getting established.
The staggering volume of subscribers is not a challenge for CSPs anymore. However the real task at hand is managing multiple layers of complexity brought on by the volume and variety of information available with them in networks, enterprise systems and partner ecosystem. Within this data deluge lies a goldmine of opportunities that Big Data Analytics manages to glean out. However, Big Data will complete only half the story if it isn’t combined with Fast Data. With mobile phones being constant companions, there are myriad opportunities that are created every moment which the CSP can capitalize on. In order to do that, constant monitoring and faster response cycles are a pre-requisite. When Big Data combines with Fast Data, those results can be achieved.
While traditional Big Data tools effectively store and analyze historical data in the “data lake,” Fast Data captures data as it streams in especially from the network.
In an offline system, data is pulled into a big data system and the analytics engine can only detect trigger once it has happened (in the past); by the time a CSP reaches out to the subscriber, it may be too late to act. In contrast, real-time decisions or recommendations that are event-triggered result in much higher response rates, because they lead to contextually relevant actions. As a result, real-time decisions improve CSP’s marketing effectiveness.
There are a number of use cases in Telecom vertical that prove the value of integrating Big Data-Fast Data streams for analytics and decisioning. For example, real-time recommendations can be configured as part of automated targeted marketing campaigns. If the subscribers meet a certain usage criteria or reach required loyalty points, a trigger can be generated to send them special offers. Similarly, when subscriber opts for a certain product, a complimentary product can be extended after making sure that she has enough in her balance to spend.
Another example could be based on events that trigger action. For example, if a subscriber enters a certain geo-location, relevant advertising offers can be sent to increase footfall to a mall or an outlet in the location. Any privacy concerns that may arise here can be eliminated if the right analytics technology is used that honor privacy and permission, in addition to anonymizing subscriber data to make it ‘Non- Personally Identifiable Information (Non- PII).
Real-time analytics with streaming data can also be very effective in arresting churn and winning back subscribers showing churn symptoms. The CSP benefit from long-term sight (Big Data), such as “this subscriber is likely to return and needs to be watched,” as well as a real-time insight, such as “this subscriber has made fewer calls today than the average calling pattern.” The combination of historical/real-time insight helps the CSP to reach out immediately with relevant offers. Event-driven triggers and learned intelligence are combined to build an offer for the subscriber that is delivered in real-time through his/her preferred channel.
Personalization of products and services based on the context could also have a significant impact on improving subscriber experience. Big Data analytics, based on historical data, informs the CSP about the inclinations and preferences of the subscriber whereas real-time data tells the CSP what the subscriber is engaged in at the present moment. When these two are combined, tailored promotions could be extended that suit the current context of the subscriber. For example, a high value subscriber is watching videos on mobile and experiences data throttling. With real-time analysis, this is detected and the subscriber is immediately upgraded to a higher speed network. This will have immediate positive impact on the subscriber’s experience, which in turn will benefit the CSP.
To understand the impact of real-time processing, let’s look at its application in a practical scenario. Flytxt offers Big Data analytics solutions to its CSP clients across globe. In one of its deployments, Flytxt processed events streaming each day from more than 200 million mobile subscribers. These numbers – from just one provider – highlight the data management challenges within the telecommunication sector, which demands more capability than traditional Big Data platforms offer. Flytxt has partnered with VoltDB, an in-memory operational database and leader in fast data, to enhance its real-time capabilities.
Using VoltDB, Flytxt processes nearly four (4) billion events every day. The results are visible in numbers. Real-time trigger campaigns yield a 40 percent to 300 percent higher conversion rate and a reduction in response time to customer events – 24 hours vs. 30 minutes – compared to non-real-time campaigns.
Big Data insights coupled with in-stream data processing empowers the CSP to remain prepared for any opportunity that may arise in the future, no matter how small is the window or how perishable is the opportunity. It allows the CSP to have a 360 degree view of the subscriber that is essential for improving subscriber experience, personalizing communication and maintaining quality of service. With faster decision cycles and time-to-market products and services, the CSP can generate significant incremental revenue.
