Key Takeaways: Application performance monitoring (APM) was built for a world that no longer exists—where applications ran in controlled environments. Cloud, APIs and the Internet are the substrate of digital experiences today, making Internet Performance Monitoring (IPM) an attractive approach to achieving the true visibility required to ensure reliability, resilience and user satisfaction.
APM Was Great, Until it Wasn’t
Application performance monitoring has been a cornerstone of IT operations for decades, tracking application health and troubleshooting performance issues. And for a long time, it has been enough. When APM was born, applications lived in a controlled environment—on-premises, usually monolithic, running on dedicated infrastructure. Even moves to off-premises cloud did not change this significantly for many enterprise applications. So if performance suffered, APM had the vision to be able to tell you why.
But that world no longer exists. Modern applications are built on cloud services, microservices and application programming interfaces (APIs); and they stretch across multiple providers, platforms, networks and geographies. Arguably, the internet is now the enterprise network. APM wasn’t designed for this; it tells you how well your backend is running but it won’t tell you why users in Singapore are experiencing slow load times while everything looks fine in Virginia.
This is something that was emphasized repeatedly in a recent conversation between Mitch Ashley, an analyst at The Futurum Group, and Mehdi Daoudi, CEO of performance management platform vendor Catchpoint. Mehdi pointed out that traditional monitoring approaches fail to account for external factors like ISP congestion, CDN slowdowns, and DNS failures—all of which can degrade user experience without triggering a single APM alert.
Addressing the Internet-Driven World
A few thoughts:
- The Internet Is Your Platform
Your users aren’t just sitting in your office. They’re accessing essential services from a web of ISPs, mobile networks, VPNs, and clouds, across the world. APM will tell you that your database query is executing efficiently, but it won’t tell you that a congested ISP peering point is causing unavailability in Central Europe.
Catchpoint’s approach is what they call Internet performance monitoring (IPM). Mehdi described the company’s IPM as founded on a global set of real-world intelligent agents that continuously track network conditions, ISP latency, peering issues, and so forth. This gives you visibility into global performance problems so you can mitigate issues beyond your own systems, uncontrolled by you. If you can’t see the entire delivery chain, you can’t fix what may currently be affecting your users.
- Third-Party Dependencies Can’t Be a Blind Spot
Modern applications rely on a messy web of APIs, SaaS integrations, and cloud services public and private. Payments, authentication, analytics, content delivery, and on and on—you know best for your applications—are handled by third parties. When one of them fails, your users don’t know or care whose fault it is. They turn to you.
As Mehdi explains it, the failure of a third-party service is still your problem—your brand suffers the damage, and your team deals with the fallout. APM provides insight into your own code and infrastructure while IPM gives you visibility into everything else—monitoring DNS resolution times, tracking API availability and latency, and catching outages and service health issues from your third-party providers before they affect you.
- Network Performance Is the Too-Often-Overlooked Failure Point
Your application’s performance is only as good as the network that delivers it. Packet loss, BGP misconfigurations, ISP throttling, access attacks beyond the VPN, and so many other issues APM can’t catch can bring an application to its knees.
IPM provides continuous visibility into the network pathways between users and your application so you know exactly where performance bottlenecks or choke points are—whether it’s an overloaded CDN node or an ISP routing issue.
IPM Can Also Improve CI/CD and Release Management
As you modernize on so many fronts, you’re pushing code faster than ever. CI/CD pipelines that are humming deploy new releases multiple times a day. But here’s the problem: if your testing and monitoring stop at the application layer, you’re not catching performance issues that arise outside your systems.
IPM can address that potential source of release snags and rollbacks. Catchpoint, for example, integrates directly into CI/CD workflows so you can detect and mitigate performance degradations just as rapidly as you release—and before users feel them. Here’s how:
- Internet-facing services are monitored as part of every deployment. Before a release, IPM can simulate real-world traffic from multiple regions, ISPs, and cloud environments to catch performance regressions early.
- Rollbacks are automated. If a release introduces slowdowns due to external dependencies—like a third-party API latency spike—IPM data can trigger automated rollback policies and maintain application service health before you might even know of the issue.
- Network resiliency is ensured as code. IPM enables organizations to treat Internet health as part of their infrastructure monitoring, embedding network performance thresholds into CI/CD pipelines.
A New Equation: APM + IPM = Full Observability
I think I have been clear enough here that I don’t think APM is “dead.” Far from it. It remains essential for tracking application health, debugging issues, and optimizing backend performance. But APM alone is like checking your car’s drivetrain and suspension while ignoring the road conditions. If the highway is full of potholes, it doesn’t matter how well the engine runs.
Pairing APM with IPM gives operations and platform engineering teams:
- End-to-end visibility, from backend code execution to real-world Internet conditions.
- Faster root cause analysis by correlating internal and external performance issues.
- A proactive, automated approach to monitoring, reducing downtime and user-experience failures that’s comprehensive, and therefore more effective.
APM was built for the previous generation of application architectures. IPM is built for the way applications run today. Your organization depends on the Internet just like everyone else’s, so you need to monitor the Internet’s effect on your applications.
As Mehdi put it, “If you don’t monitor the Internet, you’re not monitoring your business.” Anything less and you’re flying blind.
Want to learn more? Check out Catchpoint’s blog or their latest SRE Report for deeper insights into the future of monitoring.
