Groundcover’s $100M Bet: Observability Is Finally Due for a Shakeout

One hundred million dollars buys a lot of things — a modest island, a fleet of Gulfstreams, or, in Groundcover’s case, a seat at the big-kids table in cloud observability. The startup announced the nine-figure round this week, led by Sequoia with participation from Greylock and a handful of strategic angels. That kind of capital doesn’t land on a two-year-old company because investors love dashboards; it lands because the category is overdue for a reckoning.

The incumbents have gotten comfortable

Datadog, New Relic, Dynatrace, Splunk — they’ve spent the last decade turning observability into a line-item budget that CFOs dread. Pricing models based on host count, custom metrics, log ingestion volume, and “enterprise” add-ons have turned what should be a utility bill into a negotiation. I’ve watched engineering teams at three different companies spend sprint cycles just pruning dashboards to keep the invoice under six figures. That’s not observability; that’s hostage-taking.

Groundcover’s pitch is blunt: unified metrics, logs, and traces without the pricing gymnastics. They’re not the first to say it. But they’re the first in a while to back the claim with a war chest that lets them hire the kernel engineers who actually understand eBPF, OpenTelemetry, and the nightmare that is Kubernetes network plumbing.

Why now? The OpenTelemetry inflection point

OpenTelemetry reached 1.0 in 2023. That sounds like a version number, but in practice it meant every major cloud provider, framework, and instrumentation library finally agreed on a common data model. The “vendor lock-in” moat that protected the incumbents just drained. If your instrumentation speaks OTel, you can switch backends by changing an endpoint URL. Groundcover’s architecture assumes OTel-native ingestion from day one; the incumbents are still retrofitting collectors that were built for proprietary agents.

That technical delta matters. A team I advise migrated a staging cluster to an OTel-first stack last quarter. They cut agent overhead by 40% and eliminated two full-time roles dedicated to “Datadog hygiene.” Multiply that across the market and the TAM expansion becomes obvious — observability stops being a tax and starts being a platform.

The pricing trap nobody talks about

Here’s the part the pitch decks skip: storage economics. Logs are cheap until they’re not. Traces are cheap until you sample at 100%. Metrics are cheap until you hit cardinality explosions from label proliferation. Groundcover’s founding team came out of Uber’s internal observability group, where they built a columnar store that compresses high-cardinality time-series by an order of magnitude versus the incumbent row-oriented approach. If that tech transfers — and the hire of ClickHouse’s former VP of engineering suggests it does — the unit economics flip. They can price on value delivered rather than gigabytes ingested.

That’s the real threat to Datadog’s 78% gross margin. Not feature parity. Cost structure.

What the $100M actually buys

• Credibility with Fortune 500 procurement committees who equate runway with reliability.
• A sales motion that can afford proof-of-value pilots at scale — not the “two hosts free” theater.
• Time to finish the hard stuff: multi-region replication, RBAC that doesn’t require a PhD, and an API that lets you build your own UI if theirs annoys you.

Notice what’s missing: marketing blitz, feature bloat, AI-washing. Good. The category doesn’t need another “AI-powered root cause analysis” button that works on the demo dataset and hallucinates in production.

The bear case

Groundcover could still flame out. The incumbents have distribution moats — AWS Marketplace, Azure Monitor integration, GCP Operations Suite — that take years to replicate. Splunk’s enterprise accounts renew on three-year cycles. Displacing an incumbent means surviving the “we’ll match the price” conversation, which works until the CFO realizes the migration cost exceeds the savings. And if the macro tightens, $100M evaporates faster than a Series B founder expects.

But the bear case relies on inertia. Inertia breaks when the pain gets sharp enough. Kubernetes cost optimization, FinOps mandates, and the sheer fatigue of managing three observability vendors for one stack — that pain is sharp now.

My take

Groundcover won’t kill Datadog tomorrow. But they’ve forced the category’s first honest pricing conversation in a decade. Even if they stall, the next vendor negotiation you walk into will have a new reference point. That alone makes the $100M well spent.

Watch the OTel adoption curves. Watch the cardinality compression benchmarks. And for once, watch a startup that’s building infrastructure instead of wrapping someone else’s API in a prettier React dashboard. We needed this.

Source: original report