The Numbers Don’t Lie
I’ve spent enough years watching infrastructure trends cycle through the industry to recognize when something has genuinely shifted beneath the surface. OpenTelemetry is no longer the promising open source project that gets respectful nods at conferences. It’s become the default expectation, and the market numbers validate what anyone actually running modern systems already knows in their bones.
The CNCF annual survey 2025 shows OpenTelemetry now sits at 57 percent adoption among organizations running production workloads. That’s not a rounding error. That’s a supermajority. Only Kubernetes beats it in the CNCF ecosystem, and Kubernetes fundamentally changed how we deploy systems. We’re watching observability undergo the same transformation, and the proprietary platforms built on the assumption of captive customers are finally feeling the ground shift.
The financial reality is starting to show up in earnings calls and restructuring announcements. Datadog’s leadership acknowledged during Q3 2025 earnings that customer churn in the mid-market segment is accelerating, driven explicitly by organizations standardizing on OpenTelemetry. When your CFO is talking about pricing sensitivity on a public earnings call, you’re not looking at a niche trend anymore. You’re looking at a structural market change.
What Happened to the Incumbents
New Relic provides the clearest case study in how quickly the dynamics can reverse. After Francisco Partners acquired the company in 2023, there was a period where the organization seemed positioned to consolidate market share and modernize its stack. Instead, by mid-2025, they found themselves bleeding SMB customers fast enough that they fundamentally restructured pricing around consumption models. That’s not strategic innovation. That’s defensive repositioning in real time.
The mechanics are straightforward if you’ve lived through enough vendor migrations. Organizations discovered they could instrument their systems using OpenTelemetry collectors, route that telemetry to multiple backends, and get better observability outcomes than they got when locked into a single vendor’s data model. Once that pattern becomes viable at scale, the business model of premium lock-in observability platforms becomes untenable. You’re competing on functionality and support now, not on switching costs.
Datadog hasn’t faced the same acute restructuring pressure as New Relic, but the company is clearly aware the game has changed. The mid-market sensitivity to pricing that showed up in their earnings report isn’t incidental. It’s the market expressing that customers now have a viable alternative to accepting whatever price increases come down the line.
The Open Source Platforms Are Winning on Momentum
Grafana Labs’ numbers are the ones that should be keeping proprietary observability vendors awake at night. Downloads of their LGTM stack—Loki for logs, Grafana for visualization, Tempo for traces, and Mimir for metrics—grew 200 percent year-over-year in 2025. That’s not growth in a new market segment. That’s growth in the core observability market, pulled directly from what used to be proprietary platform territory.
More telling than the download numbers is what happened with enterprise support contracts. They doubled in the same period. Organizations aren’t just kicking the tires on open source anymore. They’re committing resources, running these stacks in production at scale, and paying for professional support and SLAs. That’s the pattern you see when a technology transitions from “interesting alternative” to “standard infrastructure.”
The shift comes down to this: OpenTelemetry created a genuine standard that the industry actually rallied around. When you have a standard, you stop paying for the privilege of nonconformance. You start paying for the value of the platform on top of the standard, which is a much thinner margin story.
What the 43 Percent Number Actually Means
A Forrester survey from 2025 found that enterprises running fully open source observability pipelines built on OpenTelemetry reduced their tooling costs by 43 percent over eighteen months compared to vendor lock-in implementations. Before you dismiss that as typical analyst report exaggeration, consider what you’re actually measuring. This isn’t theoretical savings. This is what organizations actually spent after moving from single-vendor platforms to open stacks.
The mechanism behind those savings is worth understanding clearly. You’re not just switching vendors. You’re removing the artificial scarcity that proprietary platforms create around data egress, storage tiers, and feature gatekeeping. An open system means you can shop around for components, use the visualization layer you prefer, and choose your storage backend based on economics and performance rather than what the vendor decided to bundle.
The other piece of those savings is organizational. When your observability stack is built on open standards, you’re not dependent on a specific vendor’s hiring decisions or product roadmap. Your team can debug problems by reading actual code. You can contribute fixes upstream. That flexibility has genuine economic value once you’re large enough to employ people who can operate systems at that level.
What Happens to Your Contracts in 2026
If you have significant Datadog or New Relic commitments rolling up for renewal in the next year, you’re entering a fundamentally different negotiating position than you had in 2023. Your vendors know that OpenTelemetry standardization has given you credible alternatives. That changes the conversation.
It doesn’t mean those platforms are going away or that you should automatically rip and replace. Datadog and New Relic have built genuinely sophisticated platforms on top of the raw observability data. Datadog’s APM and security features have real value. New Relic’s entity modeling is thoughtful and useful. The question you’re answering is whether that value justifies the price premium when OpenTelemetry has made the raw plumbing work vendor-neutral.
The most likely outcome for most mid-market and larger organizations is hybrid. You might use Datadog or New Relic for specific workloads where their specialized features justify the cost, while routing the broader telemetry through an open stack. You might use both platforms in a negotiated arrangement where you’re no longer captive to a single vendor’s quarterly price increases.
The vendors will adapt. We’ll likely see more consumption-based pricing, more transparent cost structures, and more willingness to integrate with OpenTelemetry workflows rather than insist on proprietary instrumentation. The economics of the observability market are reorganizing around the standard, not around vendor lock-in. That’s a healthier market structure for everyone except the vendors who built their margins on captive customers.
If you’re currently evaluating observability tooling or heading into contract negotiations, this is the moment to leverage that structural shift. The balance of power has genuinely moved. Understanding where your systems actually need premium features and where they’re fine with solid open infrastructure will save you significant money and complexity. I’d be interested to hear what you’re seeing in your own infrastructure decisions—the market is moving quickly enough that the ground truth in your systems matters more than any analyst report.