Intercom Completes Analyze with Monitors and Recommendations for AI Observability.

Intercom has expanded its Analyze platform into what the company calls a complete observability system for AI-powered customer support. CPO Paul Adams announced the updates at Fin Labs in Paris, introducing Monitors and Recommendations alongside the existing Insights capability.

The move addresses a growing challenge for support teams deploying AI agents: understanding what automated conversations are actually doing and how to improve them. While Intercom previously launched Monitors to help teams track AI conversation quality, this announcement positions Analyze as the hub for measuring, understanding, and optimizing the entire customer experience.

Analyze now combines three core capabilities. Insights provides visibility into customer conversation patterns and AI performance metrics. Monitors, which Intercom rolled out recently, allows teams to set quality thresholds and receive alerts when AI conversations drift outside acceptable parameters. Recommendations, the newest addition, surfaces specific actions teams can take to improve outcomes based on the data Analyze collects.

The platform fits into what Intercom describes as the Fin Flywheel: train, test, deploy, and analyze. Where competitors often treat observability as an afterthought or separate product, Intercom is building it directly into its support workflow. Teams can now spot issues in AI conversations, understand why they're happening, and get guidance on fixes without switching between multiple tools.

This matters particularly for teams scaling AI support. As resolution rates climb and human agents handle fewer routine issues, the black box problem becomes more acute. Support leaders need to know not just that conversations are happening, but whether they're actually solving customer problems. As Intercom demonstrated with Monitors, the company is focused on giving teams confidence to push AI coverage higher.

The announcement puts Intercom in more direct competition with observability-focused startups and platforms like PostHog that offer product analytics. However, Intercom's advantage is native integration with its support platform and AI agents. Teams don't need to instrument custom tracking or export data to third-party tools.

Intercom hasn't disclosed pricing for the expanded Analyze features or whether they're included in existing plans. The company also hasn't specified which metrics Recommendations will surface or how prescriptive its guidance will be. Those details will matter for teams evaluating whether Analyze can replace standalone monitoring tools.

The timing aligns with Intercom's broader AI push, including the recent launch of Apex, the company's customer service-specific AI model. By building observability directly into the platform, Intercom is betting that integrated solutions will win over best-of-breed approaches as AI support becomes table stakes.