A provider of integrated triple-play services to several million subscribers of cable TV, broadband internet, and telephony offerings, the European MSO prides itself on delivering world-class customer service. Logically, it wanted to accelerate network resolution times to continue delivering the strong customer experiences its subscriber base had come to value.
The MSO was looking for a way to identify and resolve network problems faster, particularly issues that could impact customer experiences. Its operations team was struggling to rapidly distinguish between issues caused by customer premise devices and headend equipment. This delayed the MSO’s ability to find root cause of problems and subsequently resolve the issues.
With the cost of a truck roll in Germany running about 60 to 70 euros and the handling of incoming customer service calls running about 5 to 10 euros each, the provider hoped to reduce customer service costs and improve customer satisfaction at the same time.
After a five-month proof-of-value (POV) test of Ops-IQ with actual anonymized data, hidden insights and root issues were revealed and the barriers between data silos were removed. By correlating massive amounts of disparate data and running advanced analytics on it in real-time, the MSO’s NetOps and CareOps teams learned that they could identify and troubleshoot issues faster—sometimes even before they occurred. The result was a domino effect of reduced customer calls, trouble tickets, and truck rolls; enough to save the company “seven digits” annually.
Before deploying Ops-IQ, this MSO set about to validate its real business value. They wanted to ensure that Ops-IQ would in fact help them meet their goals of greater network uptime, better customer experiences, and lower operational costs.
Accomplishing all that would require reducing the time it takes to become aware of a network issue, determine its cause, and resolve it. It’s common for repair trucks to be dispatched to customer locations, only to find that a problem’s cause lies in head-end equipment, not CPE, for example. With machine learning and analytics, however, the Ops-IQ solution promised to pinpoint root issues faster and more accurately, conserve unnecessary truck rolls, improve customer satisfaction, and reel in costs.
The company validated these claims using a real simulation that involved data from six months prior, which Guavus analyzed and aggregated into predictions. From the simulation, the MSO determined the time it would take to begin reaping the rewards was five months, says the project director. The other benefit of the simulation was to get a qualitative assessment of the product, “such as learning how well it would integrate with our existing systems,” he notes.
Guavus accomplished its on-the-mark predictions and analytics by bringing together data from all over the organization, including technical support calls, subscriber trouble tickets, and truck roll data. It correlated the data, identified commonalities, and automatically determined baselines and anomalies.
From there, the software raised alarms with specifics. For example, multiple, seemingly disparate issues all pertained to a certain CPE firmware type. With that information, a problem that would have taken days to troubleshoot, was now identified and resolved in a matter of minutes.
The MSO predicts that the acceleration of problem resolution enabled by Guavus will improve its Net Promoter Score by nearly 1 point once the solution has been in production mode for five months. This represents a significant boost to the company’s overall brand.
Multi-System Operator (MSO)
A European multiple-systems operator (MSO) needed to accelerate the discovery and resolution of network issues before they impacted customer service. The provider hoped to use automation to improve customer satisfaction—and thus its Net Promoter Score (NPS)—while also reducing operational expenses.
After a five-month proof-of-value (POV) test with and without Ops-IQ, the MSO installed Ops-IQ Live Ops software in its production environment, having proven that it would derive the following benefits:
Annual OPEX savings in the “seven digits”
4 to 7% OPEX savings attributable to fewer truck rolls
Improved customer experiences
4 to 7% reduction in customer service calls
5 to 8% reduction in trouble tickets, lightening Help Desk workloads
A significant increase in overall NPS (nearly a full point)
Proven ability to unify separate silos of big data to reveal hidden problems and anomalies.
Machine learning and intelligence allow correlation of disparate data sets (e.g. calls, tickets, and outages) to enable new insights and good decision making.
Big Data at scale: Guavus has proven its ability to analyze billions of records per day.
Closed-loop actions: Automatic actions such as call deflections and trouble ticket generation can be taken when anomalies are detected.
Rapid real-time analysis of data yields immediate results.
Subscriber-level data is correlated against network-level data to see which items actually impact customers.
Conducts real-time correlations of separate events happening in the network.
Correlates and analyzes 20 million+ time series daily (arrays of numbers indexed by time)