New Relic recently partnered with ETR once again for its annual Observability Forecast, the most robust dataset in the market tracking trends in observability and monitoring. This year, 1,700 IT decision makers and practitioners from 23 countries provided insight into their observability spending plans, outage frequency and response time, the benefits and life improvements from having an observability program, and much more.

AI a clear driver for, and focus of, observability

At the heart of this year’s report is the rise of AI-powered observability. Not only is the adoption of AI technologies the largest driver for the need for observability – shooting to the top driver this year among all options in the survey – but organizations are now racing to deploy observability to monitor their own AI programs. Monitoring AI systems has moved beyond experimentation and into production, with AI monitoring adoption growing from 42% of organizations in 2024 to 54% in 2025. Leaders cite AI-assisted troubleshooting, root cause analysis, and AI-assisted remediation actions as the most impactful capabilities. The message is clear: systems have grown too distributed and dynamic for human operators to manage alone. AI is no longer just an application to be monitored; it has become a core tool for monitoring itself.

Observability and the bottom line

The business case for observability has also never been stronger. Downtime has evolved from a purely technical problem to a board-level concern. The report found that the median cost of a high-impact outage is now $2 million per hour, or more than $33,000 for every minute a system is down. Outages often stem from network failures, third-party provider issues, or internal software changes—failures that ripple quickly through interconnected systems. In this environment, the ability to spot and contain incidents early is not just a technical advantage but a financial necessity.

The study once again found that observability investments are paying off, and those with the most robust observability coverage – known as full-stack observability – see the greatest benefits. Three-quarters of organizations in the survey report positive returns from their efforts, and nearly one in five are achieving 3x to 10x ROI. The benefits extend beyond reduced downtime. Engineers experience less alert fatigue and fewer hours spent putting out fires, freeing them to focus on more innovative projects such as AI development. Leaders, meanwhile, point to higher operational efficiency, lower security risks, and better customer experiences as clear business outcomes of improved observability practices.

Yet the report also highlights a significant observability gap. While AI adoption is accelerating, 73% of organizations still lack full-stack observability, which New Relic defines as visibility across infrastructure, applications, security, digital experience, and logs. Without it, teams are left stitching together partial insights from siloed tools and data, slowing response times and leaving blind spots that can prove costly. The difference is stark: organizations with full-stack observability experience fewer outages, detect problems faster, and cut the average cost of high-impact downtime in half compared to those without it.

Tool consolidation

One of the clearest trends is the move toward observability tool consolidation. In 2023, organizations reported using an average of six observability tools; by 2025, that number had dropped to 4.4. While tool sprawl once promised flexibility, it often created as many problems as it solved, from fragmented data to higher licensing costs. Now, more than half of organizations plan to consolidate onto unified observability platforms within the next two years. They are prioritizing vendors that can unify metrics, logs, and traces under a single pane of glass, while also delivering AI-driven insights and flexible pricing models.

As organizations continue their digital transformation journeys, the stakes are only getting higher. Accepting tool fragmentation and multimillion-dollar outages as the cost of doing business is no longer sustainable. The alternative—AI-strengthened, full-stack observability—is already delivering measurable returns in resilience, efficiency, and growth.

ETR’s ongoing research with New Relic has consistently that the future of observability is predictive, intelligent, and unified. Enterprises that embrace this shift will not only safeguard revenue and reputation but also unlock the freedom to innovate at scale. Those that hesitate risk falling behind in a world where every minute of downtime carries a multimillion-dollar price tag.

To read more about New Relic’s 2025 Observability Forecast, visit https://newrelic.com/resources/report/observability-forecast/2025