The Internet Blind Spot: Why APM Alone is Insufficient for Modern Enterprises

Mehdi Daoudi, the CEO and co-founder of Catchpoint, explains why APM is no longer sufficient for modern enterprises. This article originally appeared in Insight Jam, an enterprise IT community that enables human conversation on AI.

In the wake of recent ChatGPT outages, occurring roughly once a month since the start of the year, and with the latest incident causing significant ripple effects across industries, enterprise leaders are once again reminded of an inconvenient truth: the Internet, though invisible to most IT teams, is now critical infrastructure and an extension of their internal systems. In fact, the major outage on January 23rd overwhelmed OpenAI’s infrastructure and exposed cascading system failures, triggered by a third-party database outage and faulty health-check logic (issues that traditional APM can easily miss).

Even the most robust internal systems are vulnerable without visibility into external dependencies. That’s where Internet Performance Monitoring (IPM) comes in: it extends observability beyond the application, helping teams detect, diagnose, and respond to disruptions inside and beyond the firewall, across the broader Internet Stack before they impact end-users. 

For years, Application Performance Management (APM) tools have been the standard for tracking application health, monitoring internal code execution, server performance, and cloud environments. These tools are valuable and necessary, but they are no longer enough. As the modern digital experience increasingly relies on complex, multi-cloud architectures and third-party services consumed via APIs delivered over the public Internet, APM alone leaves a dangerous blind spot. 

Unlike internal infrastructure, the Internet is not something any single enterprise owns or controls. It’s a vast, decentralized, shared resource, and its performance can make or break your digital experience. When latency spikes, traffic gets congested, APIs are unresponsive, packets drop, or BGP routes go awry, your customers don’t care where the issue originates; they know your service isn’t working. A recent 2025 Forrester Opportunity Snapshot revealed that the average company experiences 72 Internet disruptions per month. Moreover, for 42 percent of the companies surveyed, those disruptions resulted in losses of over $500,000 in the month preceding the survey, which is over $6M annually. 

The ChatGPT outages were not isolated in cases. Every week, digital services experience degradation or downtime due to issues outside their application code. From ISP-level disruptions to DNS misconfigurations or cloud service slowdowns, these Internet-layer issues are now just as impactful as a failed server or a bad deploy. Yet, many organizations still aren’t monitoring them. 

APM excels at telling you how your app performs within your environment. It’s great to understand app events, bottlenecks in your code and infrastructure, and resource usage. But it doesn’t tell you if your CDN is down in the Northeast. It doesn’t reveal if a peering problem makes your login page fail in Germany. It doesn’t show you how your SaaS dependencies are performing in real-time. In other words, APM is blind to what’s happening beyond your firewall, where your users actually are, and the real-world experience you are delivering. 

To close this gap, enterprises must expand their observability toolkit with Internet Performance Monitoring (IPM). IPM solutions are purpose-built to monitor the internal and external dependencies that underpin every digital experience, including networks, protocols, cloud services, APIs, applications, and every internet technology between users and your services. With IPM, organizations gain real-time, proactive visibility from multiple geographies and networks to understand the real-world end-user experience. It becomes possible to quickly determine whether an issue is global or localized, whether it stems from network degradation or an application bug, and whether the problem lies upstream or downstream from your own infrastructure. 

This clarity enables teams to resolve issues faster, avoid the blame game, and maintain better SLAs with partners. It also empowers IT teams to be proactive. Instead of waiting for a user complaint to flag an issue, IPM surfaces performance degradation the moment it begins (even if those issues stem from a third-party provider). 

Today’s enterprise stack encompasses more than code and servers; it now includes APIs, clouds, and global networks. This shift means the public Internet must be treated like core infrastructure, with monitoring and accountability to match. Multi-cloud architectures, hybrid WANs, and remote work have only increased our reliance on the Internet. And as digital services continue to grow more distributed, so will the range of external variables affecting them. The only way to ensure resilience and performance is to bring the public Internet into the monitoring fold. 

Modern businesses must recognize that the Internet is not just a background utility that works magically; it’s the delivery mechanism for their brand promise. And when that mechanism falters, so does user trust, revenue, and reputation. The next major outage won’t be caused by your code. It’ll likely be caused by an issue on the Internet (one you can’t fix if you don’t see it coming). As enterprises mature in their observability strategies, APM remains essential, but it must be complemented by IPM to deliver a complete, modern view of service health. We can’t fix what we can’t see. It’s time to bring the Internet out of the blind spot and into full view through IPM.