Nobody talks about the second-order effect of AI infrastructure spending. Everyone is focused on the chipmakers, the hyperscalers, the power grids. What they are missing is the company that gets paid every time an AI model runs, fails, gets retrained, or goes rogue inside an enterprise environment.
That company is Datadog.
Here is the contradiction the market keeps misframing: DDOG is up nearly 95% year-to-date and trading near $260, yet most of the analyst community spent the first half of 2026 warning about hyperscaler competition eating its lunch. Benchmark just raised its price target to $330, noting the company is positioned to gain market share in cloud observability during AI adoption cycles, with projected 2026 revenue growth of 26.8% and a free cash flow margin of 26.4%. That is not a company losing to Amazon and Google. That is a company running faster than them.
The Quarter That Changed the Conversation
Datadog’s first-quarter 2026 revenue grew 32% year-over-year to $1.006 billion, with about 4,550 customers carrying $100,000 or more in annual recurring revenue, up from roughly 3,770 a year earlier. First time the company crossed $1 billion in a single quarter. The stock moved 31% on the day.
But here is what the headline number buried. New logo annualized bookings set an all-time record and more than doubled year over year, including large deals in observability, security, and data products — spanning wins in newer products like security, Data Observability, and Flex Logs. That is not optimization behavior. That is land-and-expand firing on all cylinders.
Slight tangent, but it matters: the bear thesis on Datadog has always been that usage-based pricing is a liability in a cost-cutting environment. And that was partly true through 2024. The real AI beneficiary thesis is simpler than most investors frame it: AI is making every company ship more code, deploy more apps, generate more telemetry. Traditional enterprises are building agents, spinning up GPU clusters, running inference, and every one of those workloads needs to be monitored. The usage-based model is not a liability anymore. It is the monetization mechanism for the AI capex supercycle.
What the Market Is Still Getting Wrong
The company’s net revenue retention rate, which measures how much existing customers expand their spending over time, has remained above 120%, demonstrating the platform’s ability to grow within accounts as observability use cases proliferate. That number is critical. NRR above 120% means Datadog is collecting more money from the same customers every year without spending an additional dollar of sales effort.
Remaining Performance Obligations, a key leading indicator of future contracted revenue, rose 51% to nearly $3.48 billion. That is not a metric you bury. That is future revenue sitting in a locked cabinet, waiting to be recognized.
The company landed two of the largest AI labs during the quarter, while its core business grew at mid-20% year-over-year. The AI-native cohort is the headline. The non-AI business accelerating into the mid-20s is the story nobody is reading correctly.
The Acquisition Nobody Has Fully Priced
On June 30, Datadog acquired Adaptive ML. The acquisition is intended to bolster Datadog’s AI research division, particularly in the areas of world models and large language model post-training. What that actually means in plain language: Datadog is building the ability to monitor, correct, and improve AI model behavior in production, not just flag when infrastructure goes down. Folding a reinforcement learning operations platform into Datadog AI Research could deepen its ability to automate root cause analysis, reduce false positives in alerting, and tune security responses in real time.
That is a different product category entirely. It is not observability. It is autonomous AI operations.
The August 6 Report Is the Next Test
Datadog guided Q2 revenue of $1.07 to $1.08 billion, up 29% to 31% year-over-year, with non-GAAP operating income of $225 to $235 million at a 21% to 22% margin. For full fiscal 2026, the company expects revenue of $4.30 to $4.34 billion, up 25% to 27% year-over-year, with non-GAAP operating income of $940 to $980 million. Both numbers look conservative given the Q1 momentum.
Management’s guidance appears conservative, with Q2 revenue guided up to $1.08 billion and full-year to $4.34 billion, both likely beatable given recent momentum. The pattern of conservative guidance followed by beats is now four quarters old.
Bull Case vs. Bear Case
The bull case: Gartner predicts the core observability platform market will reach approximately $14.2 billion by 2028. Datadog at $4 billion in annual revenue has enormous room to grow into that market, and the AI multiplier has not been fully captured in any model.
The bear case is real too. Cloud providers like Alphabet are actively integrating native observability solutions into their platforms, presenting a risk of commoditizing basic monitoring functions, while Datadog’s usage-based pricing structure has fueled proactive customer cost reductions and seat compression in a tightening macroeconomic environment. Competition from within the hyperscaler stack is not going away.
What matters is that the company’s competitive strength is reflected in its impressive gross profit margin of nearly 80%, supporting its ability to invest heavily in innovation. That margin structure is the buffer. It is what lets Datadog outspend on R&D while still generating meaningful free cash flow.
The Q2 report on August 6 will tell investors whether the 32% growth rate was a one-quarter acceleration or the beginning of a new floor. Given what happened with new logo bookings and the Adaptive ML acquisition, the evidence leans in one direction. Whether the valuation has already priced it in is the harder question — and the one worth sitting with heading into summer earnings.
