Wednesday, January 23, 2013

Reinventing the network dashboard around applications and subscribers

Over the last few years, mobile operators and vendors have turned their attention to improving the quality and reliability of the performance of networks, of handsets and of services.    Unfortunately, in that time, many existing tools and products have simply been rebranded with a “Customer Experience Management” and “Big Data” brush, aligning them, at least in marketing terms, towards helping reduce the industry’s perennial problem : customer attrition or churn which in most markets runs at an alarming 2% - 4% of subscribers per month.

In order to address customer satisfaction (which aside from pricing is often the biggest factor in churn), mobile operators are tasked with improving the quality of their network, and the quality of support provided when consumers call with a problem.   To do this, operators need information – information on the performance of services, of devices, of networks.    Now you would think that operators have a large of amount of data which could be analyzed and understood to improve services but here lies a paradox – the more complex and heterogeneous the networks become, the more management data is generated, the more difficult it is to get the right information at the right time and make it actionable.   To put it another way, operators now have way too much information, often of the wrong type, to make business decisions and improve customer satisfaction.    This is perpetuated by traditional Telco vendors who are focused on delivering yet more information from more systems, backhauling terabytes of management traffic from across the network under the illusion that if you throw enough processing power at the data, you can yield those illusive answers to subscriber woes.

Perhaps this isn't surprising when you consider where we've come from.   In the good old days of network management, you’d display a big map of your network, usually overlaid on a map of the country (or the globe) and any problems would blink, flash or alarm red.   Fix the red flashing light and life was good.    Then networks and services became more complex – think 2G,3G,4G, Wi-Fi, smartphone, machine-to-machine all with overlaid services.  To deliver a single service such as email or a mapping app or a web search requires the cooperation of many different systems, networks and companies.   The flashing red light, as a key performance indicator went away because we lost an end-to-end understanding of our services.   To date, as an industry we really haven’t really implemented a good alternative.

As it turns out then, finding the new “red light” means looking at the network from a user experience perspective – how each application is performing for each subscriber at key points across the network.   Aggregate this information in real-time and you recreate the red light map with a new paradigm – performance by application, by market, by service, by subscriber type, by revenue generated.  

Thinking through this problem, Vistapointe, a start-up focused on network insights, have figured out how to build this performance dashboard without having to backhaul and then analyze huge amounts of traffic from each node or location in the network.    Their method relies on placing Vistapointe Intelligence Packet Engines (VIPEs) at strategic points in the network where, based on location, they analyze and process only the critical information necessary to extract performance data, relaying not raw packet captures, but summarized and compressed information to the cloud where it is aggregated with information from other nodes to build a complete picture of each application and service in near real-time.     By performing much of the data analysis deep in the network, they've made the solution naturally scalable as servers in the cloud are concerned only with aggregation and not with packet processing.

The outcome of reinventing the mobile operator’s dashboard is that, for the first time, mobile operators can understand how each and every application on their network is performing – both those they own and monetize and those of partners and content providers - and when it doesn't perform, understand what went wrong and the impact to subscribers, to services and to revenue.

1 comment:

  1. Yes, Operators should find ways to move away from "pricing bits" to "pricing quality". Not to say that operators will have to drop the $ per GB kind of pricing over night. There could be a hybrid pricing options where premium services like video/voice/conference/gaming could be guaranteed for few $$ extra.

    Lots of people surely, talk about this. Question is whether operators have proper tools to implement all these.

    + Jude