Vercel Now Lets You Monitor Anything You Want

Vercel now lets developers create custom monitors, providing specific alerts on metrics like AI costs, performance, and API failure rates.

Vercel has expanded its Observability suite with a significant new feature that gives developers the power to create custom monitors. According to a recent announcement on the Vercel Blog, users can now build their own alerts based on specific queries, moving beyond predefined dashboards to a more tailored monitoring experience. This allows teams to track a wide range of signals critical to their applications. The company highlighted several practical examples, including monitoring costs from its AI Gateway, tracking Core Web Vitals and other performance metrics, watching failure rates for external APIs, or even keeping an eye on Vercel Sandbox CPU time. The update provides a much-needed layer of proactive control for the thousands of developers and businesses who rely on the platform for deploying and managing their web applications.

The new functionality works by allowing a user to define a monitor that tracks either a single metric, such as the average response time of a serverless function, or a ratio of two metrics. A common use case for a ratio-based monitor is calculating an error rate by dividing the number of errors by the total number of requests over a given period. These monitors are built on top of the existing Vercel Observability query language, meaning any data point that can be queried within the platform can now be turned into an actionable alert. This mechanism transforms Vercel's observability tools from a passive, analytical system into an active, real-time notification engine. It empowers engineering teams to define what “healthy” means for their specific application and receive immediate feedback when performance deviates from that baseline.

This feature launch is a strategic move that places Vercel in closer competition with specialized observability and application performance monitoring (APM) platforms like Datadog, New Relic, and Grafana. While it may not replace these comprehensive tools for large enterprises, it directly addresses a common workflow for many development teams: piping Vercel data into a third-party service just to configure custom alerts. By building this capability directly into its platform, Vercel is following a broader industry trend where Platform-as-a-Service (PaaS) providers are integrating more of the software development lifecycle into a single, unified offering. This strategy aims to increase platform stickiness by making it easier for developers to manage everything from deployment and hosting to monitoring and incident response within one ecosystem.

For teams building on Vercel, the practical implications are immediate and substantial. The ability to create custom monitors means they can now set up alerts that are directly tied to their business logic and user experience. An e-commerce platform, for instance, can create a high-priority alert if the failure rate of its payment gateway integration suddenly spikes. A content-heavy site can monitor a specific Core Web Vital, like Largest Contentful Paint (LCP), on its most important landing pages to ensure a good user experience. Similarly, a startup leveraging AI can set a budget alert on its Vercel AI Gateway spending to avoid unexpected costs. Looking ahead, developers should watch for Vercel to further enhance this offering with more advanced alerting options, such as anomaly detection, and deeper integrations with popular incident management tools like Slack and PagerDuty, solidifying its position as an all-in-one platform for modern web development.

This moves Vercel beyond generic observability into a highly configurable monitoring tool. For developers, it means faster, more relevant alerting on the exact metrics that define their application's health, from Core Web Vitals to specific API error rates, reducing noise and improving incident response.

Custom monitors enable businesses to proactively manage operational costs, such as AI API usage, and protect revenue by catching performance degradations faster. This direct line of sight into application health and spending allows for better resource allocation and risk management.