Insurance against myths: Device-centric vs. service-centric approach and why both are essential

Today, many systems consider assurance only at the device level, using port statistics, device status, system logs, and other infrastructure- or device-based telemetry. It is useful to understand and gain insights from a device perspective, but this information is reactive.

Similarly, the primary way to discover that a customer or end user is affected by network performance issues in this scenario is through trouble tickets. However, if a customer has already taken the time to call or create a trouble ticket, this also puts IT support in reactive mode, looking for the problem.

In this blog, we’ll compare some common methods for managing your network’s performance and reliability. You’ll learn what different assurance approaches can do to provide a holistic view of your customer or end user’s network experience, and how a more proactive, customer-centric approach can help you stay ahead of issues.

Device-level insurance is helpful, but it’s not enough

Device-level assurance is very effective at detecting severe failures and providing information that is typically red or green: red means the device is not performing as expected, and green means everything is fine. Gradual degradation is harder to detect, but from the customer or end-user perspective, any degradation slows down the network.

For example, a recent network analysis conducted by one of our service provider customers revealed that even a 0.53% packet loss can result in a 50% decrease in data or throughput. A five millisecond delay can result in a 10% decrease in throughput. Device-level insights are not effective at detecting quality of experience (QoE) issues and do not reveal the impact an issue may have on the customer. Red at the device level does not always mean the customer experience is affected, and green does not always mean everything is fine.

Service Level Assurance provides a proactive, customer-centric view

By focusing on the service, you can gain a holistic view of how the customer experiences the network and the impact of any performance delays. This allows you to take a proactive approach by continuously monitoring the end-to-end service experience.

Granular KPI measurements, such as packet loss detection below 1% and other one-way metrics, can provide precise insights into customer experiences. Small losses can have a significant impact on time- or latency-sensitive services, and a customer’s perception of slowness is measured in milliseconds.

Metadata is also a key component of service assurance. It can include a customer’s location, region, class of service, geographic coordinates, topology, or other details that add context to performance data and KPIs. This enriches knowledge and helps you understand the relationship between performance patterns. For example, if you see that all customers with a latency issue are using the same router, understanding this relationship can help you isolate the possible root cause.

Using machine learning algorithms and analytics will allow you to correlate relevant data and identify the problem. You can even get a glimpse into the future and start predicting performance by establishing a baseline of what is normal, detecting deviations, and taking preventative measures to avoid customer-impacting issues.

Bringing it all together with a unified vision of the digital experience

Correlating a single view of device-level assurance with service-centric assurance can save you time and money while helping you continuously improve the end-user digital experience. In fact, 75% of IT leaders plan to enable end-to-end visibility into all network domains from a single console, according to the Cisco 2024 Global Networking Trends Report. Organizations that do so will enable more proactive, customer-centric network operations with the ability to see, detect, and even predict customer-impacting issues instead of reactively responding to problems only when end users call and open trouble tickets.

An end-to-end view of service performance lets you focus on the customer-impacting issues that need to be prioritized. By bringing together device-centric and service-centric assurance, you better understand what the real problem is and where it’s occurring within your network.

With Cisco Provider Connectivity Assurance (formerly Accedian Skylight), enterprises gain microsecond-level visibility and service-centric insights essential for business-critical connectivity and management of large-scale, complex provider networks—the “owned” side of the global network.

The result? Simplified operations and seamless digital experiences in carrier-grade environments.

Learn more about cutting-edge approaches to insurance in our research paper: