In today’s hyper-competitive environment, operators need to be able to continually introduce new devices and services to satisfy consumers’ insatiable desire for the next shiny object — and it’s expected to come with the best customer experience. The stakes are high. It’s commonly known that customers with a shaky network experience are three times more likely to churn. The challenge historically has been that operators have no way of knowing what the full impact each new service will have on the network, the in-home WiFi experience or even on the device itself.

By applying a new breed of real time analytics to data that is collected, correlated and curated across network, operations and care, operators are able to derive insights to improve service operations and significantly enhance the customer experience. What follows are some of the ways that operators can more accurately predict the impact of new services and devices on their network by applying analytics to big data.

Visualize the Impact (Before Customer Fallout)

Current methodology around change management deployment relies on interaction between multiple teams to understand any negative impact. For example, care teams might notice a spike in calls (and let’s not forget that more than half of customers experiencing issues will never call for support). In parallel, network operations investigate the details of those customers, and ultimately tie this back to a correlated change event, such as a software update. This manual process delay often results in more customers being impacted, and more operations folks spending time on triage instead of remediation.

Streaming analytics, on the other hand, provides immediate notification of any event that causes any type of customer service disruption to the appropriate downstream system or end user to remediate the issue. This also allows the operations team to track such deployments and immediately visualize any related fallout. By tracking and visualizing deployments in real-time, a team can make necessary adjustments to the deployment such as backing out a change.

Advanced analytics can monitor all possible single and multi-attribute combinations to identify anomalies and contextualize the root cause, empowering operators with real-time information on the service effect of their change. It can also determine which devices or elements are driving abnormally high customer experience issues or care events so that operators can act swiftly to address the issue.

Put Data in Proper Context

By fusing streaming network data with stored business data, operators can see bigger trends and emerging issues before they spread to a wider footprint. This newfound context allows operators to predict, prioritize and more efficiently operate their networks. These insights provide the foundation for operational intelligence, which enables a proactive approach that prioritizes what is happening in the network before it impacts a broader set of customers. Valuable insights include things such as anomaly detection, root cause analysis and commonality correlation that predict and proactively manage quality of service and a more personalized customer experience. 

Take False Alarms Out of the Equation

The traditional troubleshooting paradigm has meant thousands of alerts and KPIs per network element. Each false alarm detracts attention from the real issues. It’s estimated that more than half of alarms received are not customer impacting, or are simply symptoms of another, potentially bigger issue.

This alarm “noise” causes delayed responses to bigger and more severe issues, which can cost upwards of $100 million per year in care interactions and churn. False alarms can be avoided by using analytics to identify anomalies by aggregating alarms related to the same issue (e.g. downstream alarms), prioritizing the events based on potential impact, and creating a detection and triage environment that identifies emerging issues before they impact the customer. For example, a global operator used analytics to predict with 93 percent accuracy which alarms would convert to customer care incidents and prioritize alarms that had the highest impact on subscribers.

Create Real-Time Deterministic Data

Operational intelligence is about balancing the speed of data with data enrichment (real-time correlation and contextualization). The goal is to analyze the data quickly and produce deterministic data, which tells operators where to put their focus. Deterministic data goes beyond a likelihood that something may or may not happen; it gives operators a specific consequence and impact for a single or series of actions.

For example, when introducing a new gateway, customers on that node will experience outages. Using these data points allow operators to look into details that can highlight the type of users that are impacted by service degradation. Perhaps the node serves a retirement community, made up of historically light Internet users. The outage may have less of an impact on this population than if a node serves college households with heavy video and Internet users. This context allows for a predictive outcome whereby, operators can assign valuable resources for the greatest customer benefit.

Summary

Operators are focused on all the components that make up the customer experience. They strive to deliver the best possible experience, while at the same time controlling the cost of care and field operations. By providing critical context and being able to predict outcomes with deterministic data, operators have an unprecedented 360 degree view into their customers’ experience, allowing them to prioritize resources, ultimately leading to proactive service assurance. Armed with newfound contextual insights, operators are able to drive costs out of their business, prioritize network fixes with a confident view on where to put valuable resources for the greatest impact on the customer experience.

Chris Menier is vice president, products and strategy at Guavus.