The New Battleground
WiFi at home has become the new battleground – literally. “Behind you, behind you! Build! Climb! Get outta there!” If you know kids anywhere from ages 8 to 18, this is sounding familiar. It’s a common occurrence in my house. I’ve got one teenager engaged in a Fortnite battle in the living room, another one completing the duo in the family room, a tween purchasing a new skin and younger son trying to stream Sponge Bob. Then the dreaded moment … I try to send a 16.2M PowerPoint presentation to my boss. The presentation gets stuck in an eternal ‘sending’ loop, Sponge Bob turns into a bunch of indistinguishable pixels and the worst thing ever happens – Fortnite freezes and the boys are down. “What happened?!?! We were number 2! We could have got the win!! Our WiFi stinks.” Reboot the router and start again.
Never-Ending Demand for More Bandwidth
At the 2018 Linux Forum Open Networking Summit, it was predicted that the average internet user will consume 1.5GB per day and connected cars and airplanes will consume 4 and 5TB per day respectively by 2020. This great flood of data brings great opportunity but also the likelihood of traffic congestion, poor download speeds, jittery streaming movies and songs, and the ultimate horror: glitchy video games. Both wireless and cable networks need to be rethought in order to handle both the high volume and dynamic swings in traffic. Case in point: Fortnite, Battle Royal – it turns out my boys are not the only ones playing. The multiplayer online game has taken over the internet by storm, jumping from 2 million concurrent players in January to 3.4 million+ concurrent players by February. The demands placed on the network are staggering and difficult to predict.
Constant Network Upgrades to Keep Up
In order to keep up, cable and internet providers are constantly upgrading their networks both at the head-end and customer premise to handle new applications and services. However, even well-planned changes can cause havoc. Gartner estimates that approximately 85% of all performance incidents can be traced to code changes, data changes, workload changes and infrastructure topology changes. Operators know that planned changes often cause incidents, but the problem becomes how to identify which changes are the ones causing problems.
Traditional Customer Care Ops Systems Lack a Wholistic View
Operators have invested substantially in Network Management Systems (NMS) and Element Management Systems (EMS) which are very good at identifying single element failures within a system. However, these systems are not as good at identifying issues impacting an end-to-end service, and most use manual thresholds to trigger alarms and call out anomalies.
Customer Care Ops teams typically monitor customer service call volumes separately. If the number of calls being received is above normal threshold volumes, one Regional Operations Center (ROC) will call other ROCs to ask what they are seeing in their area (e.g. if there any known network outages or servers down, if there have been any changes or upgrades, etc.) and manually piece together a likely cause for the higher than normal number of calls. This approach is quite time-consuming and makes it difficult to rapidly identify common attributes and probable causes.
Customer Experience is Affected
Logically, the longer it takes to identify what the true problem is, the longer it takes to fix it; and more customer service calls continue to come in. While Care Ops scramble to determine whether the cause is a head-end or other network element, or a problem with a certain type of cable modem and/or software version, the tension and the frustration of the end user customers escalates. It’s no wonder that customer satisfaction can be a constant challenge.
The annual American Customer Satisfaction Index (ACSI) Telecommunications Report polls thousands of customers across industries to discover how satisfied they are. This year the satisfaction level for the subscription TV service declined by 3.1% and Internet service followed suit. Fortune cites: “Digging into the results, cable industry customers were increasingly dissatisfied compared to 2017 with picture quality, helpfulness of staff, frequency of service interruptions, and call center assistance.”
Using New Tech to Improve Customer Experience
Cable Operators now have the opportunity to use advanced technologies, such as machine learning and analytics, to change this perception by better managing bandwidth demands to ensure high QoS, minimizing service disruptions, and remedying glitches before the customer even knows they occur. All of this leads to proactive and insightful customer service which translates to happier customers, higher customer satisfaction, higher Net Promoter Scores (NPS) and less churn. As renown Customer Experience expert Paul Greenberg, author of CRM at the Speed of Light, quotes:
“If a customer likes you and continues to like you, they will do business with you. If they don’t, they won’t.” — Paul Greenberg
Real-time Analytics Connects Care Events, Customer Calls and Network Details
Let’s go back to our original example of an Ops Center noticing a high-volume of calls. What if the Ops Center had a wholistic view of their entire network, in that precise moment? What if they could see the calls in relation to a recent upgrade – did calls spike immediately following that upgrade, indicating it may have caused an outage? Or do most of the customers calling in have the same brand and version of cable modem, with the same software running? Are these customers connected to the same head-end or CMTS? If all of this information were at the fingertips of the Ops Centers, the time to diagnose the true culprit of the problems and remedy the situation would be dramatically reduced.
Transform Customer Care Ops
Real-time analytics is transforming customer care. Using real-time data, Guavus is helping Cable Operators around the world eliminate old school manual processes and replace them with cross-silo correlation and commonalities analysis for rapid identification of real problems rather than just the symptoms. By employing automated multi-variate baselining, baselines for datasets of all types are automatically monitored and adjusted on the fly, eliminating the need for constant reprogramming of thresholds and alert levels. This enables real-time anomaly detection of true problems rather that one-off outliers, eliminating noise and shortening the time to detection of negative events. Care Ops teams are better informed and can react faster. Problems can be remedied before customers are affected and a high quality of service is retained.
Customer Satisfaction Scores Improve with Real-time Analytics
In a recent proof of value (PoV), a leading cable provider in Europe, outlines the A/B test they performed to evaluate the effectiveness of using real-time big data analytics in their network, using actual data. The team went to great pains to ensure the validity of the data and that all the data was in compliance with GDPR. After a comprehensive PoV, they were able to clearly link the use of Guavus real-time analytics to a 4-7% reduction in customer calls, a 5-8% reduction in tickets, and a significant improvement in their customer satisfaction levels, as measured by an increase in their overall brand NPS.
Improved customer sat through more uptime and better quality service? Not a problem. Now, if it were only as easy to get my boys more interested in the real-world than the virtual one, I’d be all set!
Image attributions: Fortnite Battle Royale, Epic Games and Bigstockphoto.com