It’s 2025, and while DevOps teams focus 99% of their time on their code and databases, their applications increasingly rely on complex infrastructures that span private data centers, public cloud services, APIs, SaaS tools and networks they don’t own or control. User experience of modern apps is as dependent on the reliability of the internet as it is on the performance of your code.
Yet far too many organizations are relying on legacy observability tools built for a time when everything lived behind the firewall (usually in the cloud). These tools provide excellent insight into internal systems (code traces, event and logs, and infrastructure metrics) but fall short when problems occur in the “middle mile,” last-mile ISPs, or within third-party services your customers depend on.
This is a critical blind spot, and it’s putting reliability, performance and even brand trust at risk.
The Observability Gap
The average company experiences 72 internet disruptions per month, according to a 2025 Forrester Opportunity Snapshot. Moreover, of the companies surveyed, 42% note that those disruptions resulted in losses of over $500,000 in the month preceding the survey, adding up to over $6M annually.
Observability platforms are generally focused on application code and infrastructure (MELT – metrics, events, logs and code traces), which is critical for developer teams. If something goes wrong in your code, these platforms are needed to identify and resolve issues as well as to optimize performance and resource utilization. However, in the hybrid, distributed, service-oriented digital environment we operate in today, needs are fundamentally different.
Applications are now more often a collection of dozens of interdependent components and services that require reliable, performant connectivity. As an example, an average website requires hundreds of calls to dozens of providers to load a page: DNS, SSL, content servers, images stored on object storage, fonts, advertising components, tag management, tracking, video services, CDNs, etc.
Today’s incidents, whether an outage, a slow-loading application, or an app that is not accessible to some users, are most likely not due to a bug in application code or the cloud infrastructure it runs on but due to slow DNS resolution, an unresponsive API, latency or dropped packets between cloud services, or a peering issue between two ISPs halfway across the world – the list goes on. This is where traditional APM (application performance monitoring) tools end and where internet performance monitoring (IPM) begins.
True observability requires visibility into everything that can impact user experience. You can’t afford to ignore the components that your observability platform cannot monitor, or those that are not effective at monitoring.
What IPM Brings to the Table
IPM gives DevOps teams continuous, proactive visibility into how applications and services perform from the perspective of the end user. It’s actually doing more than tracing internal requests; it’s also measuring performance across the entire delivery chain, from the cloud to the CDN to the last-mile network, from the perspective of the user. This takes advantage of a network of thousands of agents deployed around the world. This approach gives teams a more holistic, birds-eye view of what’s going on, and like a GPS tracker, it can more quickly help teams map to the heart of issues as they happen in real time.
By integrating IPM into observability stacks, teams can accomplish a lot more, including: 1) Detecting issues before customers experience them and 2) Pinpointing whether the root cause lies within your infrastructure or an external dependency. Additionally, IPM can help reduce time-to-resolution by arming on-call engineers with precise, external telemetry and help build resilience by proactively monitoring internet routing changes, DNS resolution times, and synthetic transactions across geographies.
When something’s slow, knowing why, where and whether it’s in your control to fix it has never been more readily available or convenient.
From Defensive Firefighting to Offensive Prevention
By utilizing IPM, teams can make a shift from reactive to proactive observability. Rather than waiting for a fire, teams can proactively put them out before they even get started by detecting degraded performance trends or anomalies in any location, in real time. This approach often uncovers issues that would otherwise go undetected for hours (or only surface through a spike in help desk tickets or angry social media posts), which can save companies’ brand trust and millions of dollars in the long run.
Take the example of a leading eCommerce platform that integrated IPM into its workflow. One Monday morning, their site traffic appeared to drop significantly in parts of Asia. Their APM tools showed no errors, and cloud infrastructure metrics looked healthy. But IPM alerted the team of a regional DNS issue with a third-party provider, which prevented users from resolving their site domain.
Because of the quick insight, the DevOps team successfully rerouted traffic within minutes, minimizing lost revenue and preserving customer trust. Without IPM, that issue might have gone undiagnosed for hours, or worse, attributed to internal systems, wasting precious triage time.
Building an IPM-Ready Culture
For many DevOps teams, adding IPM means completing the big picture by adding a system that relies on synthetic monitoring, BGP, RUM and many internet-centric tests to ensure complete coverage of the internet stack and the analysis tools that provide the best view into a system: Dependency maps, experience scores, data exploration smartboards, etc.
But more than tools, it’s about mindset. Embracing IPM requires shifting the focus from system KPIs (like error rates, or CPU utilization) to user-centric metrics like experience level objectives with business goals in mind. It also involves treating the internet as mission-critical infrastructure, not a black box, and ensuring that performance metrics are shared across teams, not siloed in IT or networking. Observability teams become digital operations teams, ensuring every digital interaction that makes the business work.
What Does This Mean in the Long Run
At a time when hybrid infrastructure is the norm and internet dependencies are unavoidable, achieving true observability means embracing a broader view. The best DevOps teams of 2025 are those who understand that visibility must start with the user and include everything impacting their experience: Internal and external across the internet stack.
With internet performance monitoring, teams can eliminate critical blind spots, be better prepared to manage incidents and move confidently from reaction to prevention. Because when the next performance issue hits, it may not be coming from your code, but your customers will still expect you to solve it.