Seven things I know about Network Analytics

The Mobile Network Magazine, June 2016

Network analytics is so in right now

Working in networks analytics over the last two to three years, we’ve seen the momentum of the market move towards us in a radical way. Key megatrends such as automation, virtualisation and analytics-fed orchestrators and network controllers, the rise of the IoT, and most crucially, the advent of CEM as a core strategy — with its aim of building a holistic view of the customers experience — have all driven the requirement for the delivery of real-time network analytics across the operating environment.

Companies with the ability to access some part of the network data picture have scrambled to add analytical capabilities, harnessing the capabilities of “Big Data” platforms, leading to a growing convergence between IT-based data analytics companies and network visibility vendors. This growth shows no sign of slowing down — with the network analytics market estimated by one analyst to grow from $768 million in 2015 to £2.3 billion by 2020, with communications service providers by far the largest investors in the capability.

But beware those selling network analytics as the cure-all

There’s no doubt that network analytics has truly arrived as a key capability in the network, feeding and enriching the whole gamut of functions within the CSP. But while there are indeed many business functions that can benefit from network analytics, there’s a danger that the term itself becomes overused, as if merely invoking the term can cure all the symptoms operators are trying to address. If you are using network analytics as a foundational capability to introduce more personalised services, more targeted marketing, quicker response times in customer care, more efficient network operations, or to feed network controllers and orchestrators, then it is important to understand the disciplines involved.

Developing a network analytics discipline

At Polystar we have learnt that building a network analytics discipline requires both focus and flexibility. It takes an ability to focus on the key operational targets that you want to achieve and design the data analysis regime accordingly. Yet alongside this focus you require flexibility. There’s no such thing as an apply once, use often cure-all because, although you are deriving insights from a common data source, learning what you look for and how you deliver those insights is all about flexibility. It has taken the networks analytics industry a long time to learn this lesson, and some are yet to learn it.

Data is a firehose, filter out the useful

The reason some have been slow to learn this lesson is that in many cases the application of network analytics started out in, and was closely tied to, the network and technical departments within an operator. These are the functions used to working with large network management and monitoring systems, requesting and analysing thousands of KPIs to help them assess network performance. So the analytics tools were designed to map the existing lists of KPIs, and there were thousands of them, that the network ops teams had asked for.

Networks analytics developers were creating views of the data that were predicated on the discovery and analysis on thousands of data sets that were of little relevance to, and made little sense to, other business units within the operator. This slowed down development, integration and implementation, and made them operationally complex. Focus and flexibility filters out the useful from the data firehouse, delivering only relevant information to business units.

Robert-Eriksson-polystar

Finally, I know that many operators have already made the first steps to embed the benefits of network analytics across their business. These operators are already building a competitive advantage as they face the future.

It’s about who knows what, not just what you know

Once you have liberated network analytics from this technically-focused strait jacket, you truly start to see the benefits of democratising data across an operator. Democratisation of data means giving a much broader range of functions and people access to data based on networks analytics. This can mean your customer care team being able to understand a customer issue much quicker, or you marketing team being able to structure offers based on network usage and performance, or your C-suite gaining a real time but condensed view of key revenue-impacting events in the network.

It’s as much people as process

And once that happens, it quickly becomes clear that the biggest change catalyst for extracting the benefits from a flexible network analytics platform are the people within an operator itself. Once staff members are exposed to data sets that make sense to them, then they want to do more, and get more creative in their use of network analytics-based data. You start to both feed and harness the desire of your staff to do a better job for their customers. It’s a virtuous loop.

The future of the network is analytics-fed and analytics-led

If you look at the direction of change in our industry, whether at the network layer or in the transformation of network operators into digital service providers, the foundational layer of everything is analytics. Automated, programmable networks rest upon a new level of network, service and application visibility that, in turn, requires network analytics embedded into the edge and core of the network. Giving new IoT and subscriber use cases an outstanding experience requires sophisticated analytics capabilities at the service layer. Transforming the operator into being a customer-centric business requires the delivery of data across the organisation.

 

This article is published in The Mobile Network Magazine, June 2016.

 

 

POLYSTAR MEDIA CONTACT

Inna Ott
Director of Marketing
Phone: +46 8 50 600 600
Email: inna.ott@polystar.com

Download