Datadog评级下调：模型上调但战术性降级，关注Q3+盈利预期

Peter Weed+1 917 344 8390peter.weed@bernsteinsg.com RatingMarket-Perform(OutperformOLD) Price Target DDOG 226.00 USD(180.00OLD) Datadog (DDOG): raising our model, but tactical downgrade Datadog continues to impress us with their speed and completeness of their vision aroundboth observability and the broader requirements for a VP of Product Operations. Their recentDASH conference made us incremental bullish about their AI for Datadog / Datadog for AIstory and line of sight for areas such as cybersecurity — they could end up becoming anotherlarge revenue pillar. Overall we reiterate our optimism that Datadog is both an AI winner andhas decreasing risk from disruption as new technology horizons emerge (e.g., “Born in AIheadless”?).We raise our model’s ex-AI growth duration a few % points. We raise our Born-in-AI revenue expectations (particularly Q4+).Our model post Q1implied the future “Born-in-AI” growth (outside their largest customer) after this year wouldslow to effectively ~40% YoY growth. But with the newest AI Lab customers (added in Q4and Q1) this seems much too low. While there probably isn’t a 1:1 relationship with the AI Labrevenue and Datadog (e.g., pricing increases drive AI labs, Datadog tends to discount morewith scale), we anticipate at most price is ~½ the lab’s growth. Based on the labs' expectationof 100%+ revenue CAGR over the next few years, we raise Datadog’s participation. We willneed to refine this over time as we learn more, but this feels like a more fair expectation. DOWNGRADE on Q3+ earnings caution vs. more exuberant investor expectations:demand signals slowing in both enterprise and some AI Labs.Not only do we startlapping tough comps in Q4, but as we’ve discussed in several other notes, we are seeingdemand signals flat lining ex-AI (~85% of revenue) that causes ex-AI growth to peak in Q3,and potentially regress -100-200bps in Q4. We also re-highlight signals AI Labs are seeinggrowth plateau — our QoQ into Q3 ”Born-in-AI” expectations are low relative to recentgrowth trends. All together, Q4 could see -500bps growth rate regression, falling to ~29%YoY. This would be a disappointment vs. recent investor conversations where some expecthigh 30% to 40%+ peak growth, and 30%+ growth into next year. Investment Implications Raised modeled Datadog revenue growth a few hundred bps over the next decade. Applyingour 50/50 multiples (higher ~16x NTM Revenue vs. 14x before), and DCF (11% WACC, 3%terminal growth) we raise our PT to $226 and downgrade to Market-Perform. DETAILS BORN-IN-AI MODEL EXPECTATIONS (~15% OF Q2 REVENUE) “Largest AI customer” (OpenIA?) rationalizes $170MM ARR, but remaining base grows 65% this year (deceleratingto 45% and 36% in +1 and +2 years).[we believe it is OpenAI, so use them in our model discussion here] Previously we havediscussed our model where OpenAI would rationalize ~$170MM ARR in Q3 off a contracted $210MM+ ARR. Specifically weanticipate they will move “Business metrics” to Chronosphere and Logs to Clickhouse. But the rest of the platform, ~$40MMARR estimated, should still grow with the underlying scale of the company. How fast should we grow that remaining $40MM ARR? Relative to Q3 last year, when they implemented their prior contract,we thought it seemed likely that OpenAI would grow ARR at least 200%. Even after our previously noted signals of slowinggrowth (Exhibit 1) that still seems possible (or near it) given they’d already grown ~100% by February/March. With that said, webelieve a material portion of their growth this year will be from increased pricing and ads [NOTE: OpenAI has reportedly talkedabout massive price cuts to compete even more heavily for token share — we will watch how this impacts the setup / demand /scale as the reality evolves]. The company discussed hitting $2.5B in advertising revenue this year as recently as April, and we’llassume this is attainable within the growth noted above. And from a pricing standpoint we have industry benchmarks from the“Token Price Index” that suggest real world prices are up ~85% since the contract went into place. To further validated this, wewatched revenue per engagement on their web interface as a proxy, and noticed that also skyrocketed into Q1 (Exhibit 2) — wethink this does miss revenue from Codex and their API, so is not a complete revenue picture. So, while we anticipated overall YoYgrowth to be high for OpenAI, the effective scale growth would be lower from a Datadog revenue standpoint. Based on our pricing, ads and demand flattening estimates, we get a mid-point underlying demand growth to Datadog of 65%this year. We use that to scale up the remaining ~$40MM ARR by Q4 of this year. There is some upside potential if demandrecovers or more of OpenAI's ARR growth proven to be driven by usage instead of pricing (if they bring prices down, forinstance). This is something we will need to watch and improve with our read-through signal. We will assume the incremental added this year