The Short Version: Datadog followed its billion-dollar Q1 with another quarter of faster growth. Q2 revenue reached $1.12 billion, up 36% year-over-year. The AI observability thesis has more evidence behind it, but the gap between GAAP and adjusted profitability still deserves attention. The price paid and the cost of that growth still matter.
Key Takeaways
- Q2 2026 revenue grew 36% to $1.12 billion; the full-year revenue outlook is $4.45-$4.47 billion.
- The latest reported large-customer count is about 4,720 with $100,000 or more in annual run-rate revenue (ARR).
- Non-GAAP operating margin was 23%; GAAP operating margin rounded to 0%. Read the reconciliation, not just the adjusted headline.
- The next financial checkpoint is Q3, not the already-reported Q2.
- Q1 product launches remain relevant, but do not measure the revenue contribution of AI on their own.
- No current share price or valuation multiple is assumed in this analysis.
The Layer Nobody Talks About Until Something Breaks
Every conversation about AI infrastructure starts with chips. Nvidia, AMD, the semiconductor race. Then it moves to networking – Arista, InfiniBand, who’s building the spine. Then data centers, power contracts, cooling systems.
Observability comes last in that conversation. It is also the layer that helps teams see whether everything above it actually works in production.
Here is the problem nobody discusses until it happens to them. You can build the most sophisticated AI cluster in the world – the best GPUs, the fastest networking, the most efficient data center – and still have your AI deployment fail in production because something went wrong and nobody could see it happening. A model started returning degraded outputs. Latency spiked in one region. A security vulnerability opened up in an agentic workflow. The system kept running, but it was running wrong. A billion-dollar GPU cluster can still have a very expensive bad day.
Observability software helps teams catch those problems by collecting and connecting telemetry such as metrics, logs and traces. Coverage depends on instrumentation and configuration; blind spots can remain even when the dashboard looks reassuring. In AI environments, that visibility matters, but it still has to work alongside testing, security controls and human judgment.
Datadog is an established observability and security provider. The investment thesis is that more complex AI deployments can create more demand for its platform. How much of that opportunity is already in the share price is a separate valuation question.
Experience Transparency
I track Datadog differently than I track Palantir or Arista. The thesis here is not about a single killer metric like the Rule of 40. It is about revenue growth acceleration combined with large customer expansion. When those two move in the same direction simultaneously – as they did in Q1 2026 – the operating trend is encouraging, although it does not guarantee durable stock returns. That is what I watch every quarter.
The Q2 2026 Numbers: Growth Improved, But Read Both Profit Measures
The August 6 Q2 release reported revenue of $1.12 billion, up 36%, and non-GAAP diluted EPS of $0.65. Operating cash flow was $316 million and free cash flow was $279 million. Q2 is the newest published quarter on the company’s results index as checked October 20.
For historical context, the May 7 Q1 release reported $1.006 billion of revenue and 32% growth. Q2 therefore extends the acceleration beyond that first quarter. That is a better foundation for the thesis than one strong report alone.
The Q2 financial supplement shows $257.031 million of non-GAAP operating income versus $5.455 million under GAAP. The corresponding margins round to 23% and 0%. Adjusted income is not the same thing as cash flow, and the GAAP result is not something to sweep under the rug.
The Updated Guidance: A Forecast, Not Money Already Earned
The August outlook puts full-year 2026 revenue at $4.45-$4.47 billion and non-GAAP diluted EPS at $2.50-$2.54. Q3 revenue guidance is $1.135-$1.145 billion, with non-GAAP operating income of $260-$270 million. These are management forecasts issued August 6, not reported results.
May’s full-year revenue range was $4.30-$4.34 billion. The new midpoint is $140 million higher. The raise is encouraging. Delivery now depends on customer demand and spending decisions, including how aggressively customers optimize usage and budgets.
The Large Customer Metric: Why 4,720 Matters
The Q2 10-Q reports about 4,720 customers with ARR of $100,000 or more, versus 3,850 a year earlier. These customers represented 91% of ARR. ARR here means annual run-rate revenue; it is an operating metric, not booked annual GAAP revenue or a guaranteed future sale.
Large accounts give Datadog room to expand across infrastructure monitoring, application performance monitoring, logs and security. Additional product adoption can deepen a customer relationship. Expansion still has to earn its place in the budget: customers can cut usage, negotiate pricing or choose competing tools.
Follow retained spend and cash generation alongside customer additions. Together, they give a better view of whether those larger relationships are becoming more valuable.
The Product Story: Keep the Capabilities Straight
Datadog’s Q1 release announced general availability of GPU Monitoring, MCP Server, Bits AI Security Agent and Experiments. These were Q1 launches, not a complete list of everything introduced in 2026.
GPU Monitoring connects GPU fleet health, cost and performance with the workloads consuming those resources. That is useful when expensive hardware is idle, oversubscribed or struggling. The appeal is connecting GPU telemetry with the rest of the operational picture in one platform.
The MCP Server gives AI agents and coding environments access to Datadog observability data, subject to access controls. That direction matters: agents can query operational context from Datadog. The server is not, by itself, a universal audit of every external action an agent takes.
The current Bits Security Analyst product page describes automated alert investigations and recommended verdicts for security teams. Human review still matters. Experiments supports product experimentation and A/B-test analysis with observability guardrails; it is broader than AI-model deployment alone.
The Q2 release also reported the Adaptive ML acquisition and further DASH launches, including AI Guard and Bits Agent Builder. The launch list shows where Datadog is investing. Evidence of paid adoption and revenue contribution is the next step.
Wall Street Reality Check
The business can improve while the stock becomes less attractive, or the reverse. A May earnings-day move cannot tell you whether the October entry price is reasonable. Check a dated quote, share count and valuation against more than one growth scenario. The spreadsheet still gets a vote, even when the earnings call deserves applause.
The Competitive Landscape: Plenty of Companies Want This Budget
Datadog’s Q2 filing identifies Cisco, New Relic and Dynatrace in application performance monitoring, Cisco and Elastic in log management, and native AWS, Google Cloud and Microsoft Azure offerings in cloud monitoring. It also warns about pricing pressure and changing technology.
The bull case favors a platform that connects data across tools and environments. The bear case is that some customers decide a narrower tool or their cloud provider’s offering is good enough. Datadog and those alternatives are competing for the same customer budget.
Product breadth can help retention, but it also increases the execution burden. Watch whether Datadog can win more workloads without sacrificing margins or making its platform harder to manage.
The Agentic AI Opportunity: More Complexity, More Questions
The observability thesis is straightforward: software that takes actions needs scrutiny as well as uptime. Teams need to understand what an agent did, which tools it used, what it cost and whether its output met the intended standard.
That can create demand for telemetry and investigation tools. It can also create more competition, including open-source and cloud-native alternatives. Observability does not replace permission controls, security testing or a human accountable for consequential actions.
Datadog has products addressing parts of that workflow. Paid adoption and long-term per-share economics will determine how much that opportunity is worth to shareholders.
Risk Factors: What Can Disrupt the Thesis
The GAAP-to-adjusted gap is the first one. Q2 stock-based compensation was $220.251 million, according to the financial supplement. It is a real compensation cost with potential dilution consequences. The operating-income reconciliation also includes payroll taxes on employee stock transactions, acquired-intangible amortization and M&A costs; the entire gap is not stock-based compensation.
Using rounded figures, revenue growth plus non-GAAP operating margin gives a Rule of 40 score of about 59: 36 + 23. That is a screening calculation using an adjusted operating-margin definition, not a valuation model or a guarantee of returns. Different margin definitions give different scores.
Competition, customer spending discipline and the cost of developing new products can all weaken the thesis. Even if the operating story holds up, paying too much can still produce a disappointing stock return.
What to Watch in Q3
The next report needs to be compared with the August Q3 outlook, not with an expired Q2 forecast. Four checks matter.
Revenue: compare the result and updated full-year outlook with management’s forecast, and ask what drove any change in demand or usage.
Large customers: track net additions and retention alongside the latest Q2 base, using the same ARR definition.
AI adoption: look for evidence of paid usage, customer expansion and practical deployment rather than treating product announcements as revenue.
Profit and cash: compare GAAP and non-GAAP margins, stock compensation, operating cash flow and free cash flow. Growth is more useful to shareholders when the per-share economics improve too.
Bottom Line
Q2 strengthens the operating case for Datadog. The next checks are customer demand, the economics after stock compensation, and the price paid for future growth. AI observability is a credible opportunity; durable cash generation per share is how the investment case earns its keep.
Further Reading
- AI Infrastructure Stocks 2026: The Full Picks-and-Shovels Playbook – where Datadog fits in the broader five-layer stack
- Palantir Stock (PLTR): Results and Valuation Risks – the operational AI software layer that sits alongside observability
- Louis Navellier’s Growth Investor Review – a separate review of the paid research service, not evidence for the Datadog thesis
Disclosure: This article is for informational and educational purposes only and does not constitute investment advice. Always conduct your own research before making investment decisions.