Q1'26: AI（及核心）持续爆发！

+1 917 344 8342luwei.yang@bernsteinsg.com Rating Outperform Price Target DDOG 180.00 USD(167.00OLD) Datadog (DDOG) Q1'26: Born-in-AI (and core) continues to rip! Datadog’s Q1’26 earnings delivered its largest Q1 beat since COVID, at +5.1% (midpoint),and then doubled down with the largest Q2 QoQ guide since ‘21 and largest FY raise(+5.9%) since Q1’22. We impute that Born-in-AI added around +$15MM revenue QoQ(modestly accelerating YoY vs. a tough comp), and impressively added near +$200MMARR. On their core customers, we impute growth continued to track our web metric / AWSbellwether work, and accelerated to near 26% YoY (company reported “mid-20%”). Where is all this demand coming from? How we read the tea leaves.First, they havecertainly showed off some major wins in large Born-in-AI customers. Between Anthropica quarter ago, and two of the largest hyperscalers’ AI labs, they are stacking up tens if notnearly $100MM ARR right there. But we understand (and like to see) it goes much deeper,and can be seen in the accelerating $100K+ category where net new was +240 QoQ (thelargest Q1 since ‘22). We think the acceleration in the core and in Born-in-AI are linked — the ...it may be in the back of your mind: churn risk? Not now.Yes, OpenAI seems likely toresize their usage of Datadog later this year (management remain cautious relative to their“largest customer”). But we don’t hear indications that Datadog’s other large adopters arethe same use case as OpenAI and may move elsewhere (high cardinality business metrics).So that is a positive. But we do note that the CEO emphasized current demand reflectsurgency and focus for these customers on building their own AI products, thus Datadog is agodsend to deal with the operational tooling. In the further future, as these businesses slow, Investment Implications We add a near term $200MM ARR bump in Born-in-AI customers. Little change otherwise.Applying our 50/50 multiples (higher ~14x NTM Revenue vs. 13.5x before), and DCF (11%WACC, 3% terminal growth) we raise our PT to $180 and maintain an Outperform rating. DETAILS MANAGEMENT CALLBACK AI-native customers continue to expand in number, scale, and product adoption, reinforcing platform stickiness.Management highlighted that among the 22 AI-native customers with $1M+ ARR, nearly all of them use 10 or more Datadog products, with all but one adopting all three observability pillars. This underscores that AI-native companies behave similarlyto other large scale cloud customers: they start with infrastructure, logs, and APM, then expand opportunistically across theproduct portfolio. While individual product mix may vary by architecture, the takeaway is that AI-native customers drive deep Among the non-AI native customers, AI-driven demand is manifesting as broad-based cloud consumption growthrather than a discrete, quantifiable revenue line.Management reiterated that while AI activity is clearly increasing acrossthe platform, supported by higher AI assistant usage, SRE investigations, and other internal signals, it’s not possible to directlyattribute a specific portion of core growth to AI. AI workloads manifest through higher usage of compute, storage, networking, Training workloads are emerging as an incremental observability use case, though inferencing remains the dominantdriver today.Management noted early signs that customers are beginning to use Datadog during model training, a phasehistorically underserved due to its episodic and non-production nature. The shift is driven by compressed training windows,expensive engineering resources, and the need for real-time visibility during competitive model development cycles. That said, Accelerating consumption trends appear broad-based rather than customer or product specific.Managementcharacterized recent acceleration as coming from a mix of existing customers ramping usage and new customers comingonline, rather than isolated cases of customers materially exceeding budgets. While AI-related token usage and faster cloud Public sector and FedRAMP represent a long-dated but increasingly accessible opportunity.Management reiteratedenthusiasm for the FedRAMP authorization, framing it as enabling access to US federal agencies as well as commercialcustomers that service government entities. Public sector revenue was under 1% in February 2024 and remains in a similar Sales capacity expansion continues methodically, with no near-term capacity constraints from infrastructure orGPUs.The company continues to invest in sales hiring while balancing profitability, noting improving productivity from priorinvestments. On the infrastructure side, Datadog indicated that its internally developed models are still in testing and research EXHIBIT 2:Based on the company’s disclosure of non-AIrevenue grew mid-20% in Q1, we estimate the revenuefrom “born-in-AI” companies had another strong EXHIBIT 3:The company disclosed that total NRRimproved to low-120% (we estimate it was clos