Geo-analytics: the antidote to network subscriber churn

Author: Gavin Hayhurst, Head of Product Marketing

What is the typical end-user’s reaction when experiencing poor quality of service from their mobile network? Not wanting to go through the hassle of switching network providers, one can assume that contacting the operator’s customer service and logging the complaint would be the first port of call. From the operator’s perspective, this is the most valuable path of action, as information regarding the quality of the network – as well as when and where these issues arise – help inform optimization engineers tasked with continually improving the network and delivering on key performance indicator (KPI) targets.

In practice, however, data provided by customers is rarely relied upon for network optimization. For one, operators can instead glean network quality information by analysing cell performance counters. These systems highlight dropped calls and other issues with the intention of identifying and rectifying issues before the customer has a chance to notice, or indeed complain, about them.

Unfortunately, network performance counters often don’t tell the complete story. It is quite common for all counters to look good, while certain, possibly VIP, subscribers are still frustrated by poor performance. Sadly, it is also possible for an engineer to ‘tune’ a network to meet KPIs in order to earn remuneration bonuses, while making the actual experience worse for customers.

Using customer data of course comes with its own perceived problems: reliability (i.e. customers reporting faults that are in fact due to user error); accuracy (i.e. differences in how customers are reporting issues); and validity (i.e. the number of complaints required before action can be taken). But these problems are not without a solution.

Geo-located subscriber experience data helps to close the gap between customers encountering quality of experience issues and the network engineers tasked with addressing them.

The benefits of geo-location

Geo-located subscriber experience data provides highly precise information regarding network experience in a given area, from high level analysis, right down to data from individual customer devices. This granular detail can help engineers move away from unreliable, invalid or inaccurate subscriber experience data, enabling them to identify the root cause of issues much more efficiently. When it comes to maintaining Quality of Experience (QoE), there are several ways in which geo-analytics becomes particularly useful for network engineers.

First, with network monitoring. It is vital that optimization engineers are able to spot, at a glance, if something isn’t right. While performance counters may tell one story, it may not always be accurate. This is where geo-located data comes in, providing additional value in the form of heatmaps, helping network engineers to immediately identify if, and where, in the network there are issues, and what these look like from a KPI point of view.

Second, geo-located data can be extremely powerful as far as continuous improvement is concerned. The role of a network engineer isn’t to simply ensure the network “just” works, it’s also to constantly improve the network. Often, optimization engineers will locate the worst-performing sites and attempt to improve their performance. Eventually, as each worst-performing site is improved, the overall performance across the entire network is lifted, therefore improving network level KIPs. Geo-located data becomes important when it comes to helping optimization engineers determine the changes they should make across the network. For example, by identifying where calls are being dropped, an engineer can determine the best course of action—is the issue happening close to the cell site, at the cell border, or in an area where the cell is out of its coverage area? This information can be extremely important in determining the best fix, and can prove to be much more valuable than only using cell statistics which don’t provide as detailed a picture.

Third, is investigation. Geo-located data is not only powerful in helping to detect a problem, it is also valuable when it comes to identifying the root cause of an issue. With geo-located data, optimization engineers can identify problems at a high level and then drill down, all the way to Layer 3 messages from individual calls. This eases the process of investigation, and allows engineers to resolve issues more efficiently.

Fourth, and finally, geo-located data is critical when dealing with customer complaints. This is probably the least favourite part of any optimization engineers’ job and typically involves them having to visit a disgruntled customer to recreate their negative experience on a test phone. This is often easier said than done however as reproducing poor performance isn’t always guaranteed. By leveraging geo-location, engineers can investigate the actual call during which the issue occurred. In doing so, they’ll be able to see where the problems reside, without the need to reproduce them—at the risk of failing to do so. This makes finding a resolution that much easier and quicker.

Delivering on network quality promises

Every mobile operator across the world has one common goal: to deliver an unrivalled subscriber experience. What may seem like a “simple” goal can be extremely difficult to achieve and even harder to maintain. As subscribers expect more and more from their network experience and become less forgiving when it comes to faults, outages and performance issues, it’s critical that mobile operators arm themselves with the right tools to optimize, monitor and manage network performance so as to consistently deliver the best possible QoE. Geo-located data is one such tool which can go a long way towards delivering on the network quality promises of tomorrow.