The most important earnings report this week is not the largest. It belongs to Datadog, which reports Q3 2026 results on November 4 before the U.S. financial markets open, carrying a question that every investment committee in enterprise software is trying to answer: are AI applications actually running at scale, and is anyone paying for the infrastructure underneath them?
This is not a Datadog-specific question. It is the question. Datadog faces a broad software-sector selloff tied to concerns about hyperscaler artificial-intelligence spending and changing software valuations. DDOG dropped 11.50% in the October 30 session alone. The stock has become a proxy for the whole debate.
Why Observability Is the Honest Signal
Most AI revenue discussions involve committed spend: cloud contracts, seat licenses, annual deals that show up in RPO before a single workload runs. Datadog’s consumption billing is different. Customers pay for what they actually use. When an AI model runs in production, the telemetry flows through Datadog’s platform and the meter turns. There is no accounting smoothing, no deferred obligation that flatters the look of a transition. The revenue shows up when the workload does.
That is why the consumption line matters more than the headline EPS figure. Rosenblatt anticipates Q3 results will benefit from app modernization and new enterprise AI workloads, as well as contributions from more than 750 AI-native customers, including 10 of the top 10 AI leaders. But the bull case is not about those named accounts. Management said AI is becoming a real secular growth driver, with strong traction in both AI-native customers and traditional customers adopting AI-related workloads, with non-AI customer revenue growing in the mid-20s percent.
The Bull Case and the Reservation Inside It
In Q2, Datadog’s 11% quarter-over-quarter revenue growth was the highest since Q2 2022, and the $115 million quarter-over-quarter revenue added was a record by a significant margin. Datadog raised its full-year 2026 revenue guidance to $4.45 billion to $4.47 billion and set non-GAAP EPS guidance of $2.50 to $2.54, signaling continued confidence in demand for its expanding AI-oriented platform capabilities.
The consensus going into Wednesday calls for $1.14 billion in revenue and $0.64 earnings per share. TD Cowen is predicting strong results and anticipating growth of approximately 35%, surpassing the company’s own guidance of 29%. RBC Capital has also raised its revenue growth expectations for Q3 to 34% to 35%, lifting its price target to $320.
The reservation is valuation. Datadog’s P/S ratio was about 25.2x in late October, sitting against the US Software industry at about 3.9x and peers around 8x. That premium signals meaningfully higher valuation risk if sentiment cools. Analysts at roughly 47 firms rate the stock Strong Buy. That kind of consensus is also a crowding risk.
What Investors Are Missing
Datadog’s own State of AI Engineering 2026 report found that operational complexity, not model intelligence, is becoming the primary barrier to reliable AI at scale, with around 5% of AI model requests failing in production. That finding is quietly the most important commercial argument in the company’s filing. Failure rates generate monitoring demand. Every AI application that fails in production is a new Datadog workload. The larger and messier the AI deployment wave, the more essential the observability layer becomes.
Compare that dynamic to what ServiceNow and Snowflake are selling. ServiceNow provides one of the clearest examples of AI being monetized inside an established enterprise software platform, with its AI business crossing $1 billion in annual contract value in Q2 2026. Snowflake’s AI story is driven by its lock on enterprise crown-jewel data: AI workloads increase data consumption and enterprise database spending. Those are real businesses. But both are downstream of the question Datadog answers: is the workload running at all?
Stocks to Watch
Datadog (DDOG) is the primary read. If consumption revenue accelerates above the guidance band of $1.135 to $1.145 billion, the sector argument shifts materially toward AI monetization being real. A miss reopens the debate entirely.
ServiceNow (NOW) already reported a strong quarter, but its AI contract commitments are structured differently. Datadog’s consumption data either validates or complicates the forward revenue assumptions baked into NOW’s multiple.
Snowflake (SNOW) faces the same read-through. Snowflake announced a definitive agreement to acquire Observe, an AI-powered observability company, a direct competitive signal that consumption-based AI monitoring is strategically valuable enough to buy.
Microsoft (MSFT) is the context provider. Azure AI consumption growth, disclosed in Microsoft’s own earnings, set the macro frame. If Datadog’s numbers confirm AI workloads are scaling in production, it retroactively supports the Azure AI growth Microsoft reported. If they don’t, that growth requires a second look.
