Bernstein believes Datadog is a strong company, but says the stock already prices in most of the good news expected over the next 12 to 18 months. That view sits at the center of its latest call: Datadog’s second-quarter surge in 2026 may have marked a near-term growth peak, with a mild deceleration likely in the second half.
According to a TechFlowPost write-up of a Bernstein report dated Aug. 6, 2026, Datadog generated $1.121 billion in revenue in the second quarter, up 35.6% from a year earlier. Growth accelerated by 340 basis points sequentially, the company’s strongest rate since 2022. Revenue came in $46 million above guidance, and the midpoint of full-year guidance was raised by more than three times. Non-AI revenue grew more than 20% year over year, while AI-native customers continued to contribute a larger share of growth.
Bernstein’s read is that the quarter’s strength itself may be the warning sign. The firm says the company’s own outlook already implies slower growth in the back half of the year. Even if management continues its recent pattern of beating expectations and raising guidance in coming quarters, Bernstein argues growth is more likely to hold around current levels than move higher from here. Its estimate calls for second-half growth of roughly 33%, about 200 basis points below Q2.
Strong Q2 numbers, but Bernstein sees a turning point
On the surface, Datadog’s Q2 print was hard to fault. Revenue rose 35.6% year over year, and the 340-basis-point sequential acceleration was the fastest since 2022. The quarter beat guidance by $46 million, and the midpoint of the full-year outlook moved up by more than three times.
Bernstein’s point is that such a strong print does not automatically mean another step up in the growth curve is ahead. In its view, the quarter looks more like the high-water mark. The report says the back-half setup already points to moderation, and in a stock carrying elevated expectations, even continued outperformance may not be enough to produce another round of acceleration.
That leaves the market with a narrower question: whether this is simply a moderate slowdown or the beginning of a more durable cooling trend. Bernstein leans toward the former, but does not think current valuation leaves much room for investors to absorb that transition comfortably.
Largest AI customer is seen using less, with a one-time NRR dip expected in Q3
The clearest signal, in Bernstein’s view, comes from Datadog’s largest AI customer. The firm believes that customer is OpenAI.
The report says OpenAI renewed a nine-figure contract early in Q3, but spending levels came in below prior assumptions. Bernstein now models annual contract value, or ACV, for that customer at about $115 million. That is slightly above the roughly $100 million assumed in its earlier model, but actual usage had already pulled back at the start of Q3.
Management also said on the earnings call that the customer had been fully de-risked in Q4 guidance. In practical terms, Bernstein says the company’s Q4 outlook includes only the contractual minimum spend and assumes no usage above that floor.
That setup leads Bernstein to expect a one-time dip in net revenue retention, or NRR, during Q3. The report notes that NRR had been improving steadily since bottoming in Q2 2023, but lower usage from OpenAI is expected to interrupt that trend.
For the second-largest AI customer, Bernstein believes the account is Anthropic. It expects that customer to grow 10% to 15% sequentially in both Q3 and Q4, with no major model release in the near term. Other AI customers are growing at well above 100%, according to the report, but the base remains too small to offset weaker spending from the top account.
AWS web traffic data points to a possible Q4 air pocket
Beyond AI demand, Bernstein also used AWS SSO web traffic as a leading indicator for non-AI customer activity.
The report says web traffic looked relatively soft toward the end of Q2, leaving Q3 with a weaker starting point. If weekly patterns in Q3 follow the same shape seen in 2025, year-over-year growth could face pressure. At the same time, Bernstein says a strong enough sequential trend in Q3 could limit the issue to a single-quarter air pocket rather than signal a broader change in demand.
That distinction matters to the firm’s longer-term view. Bernstein explicitly characterizes the potential Q4 weakness as a one-time issue, not a structural problem. It expects Datadog to return to its recent trend by the first quarter of 2027 and sees net new ARR growth reaccelerating to the low- to mid-teens in 2027 and 2028.
Valuation and crowded positioning are the main concerns
Bernstein maintained its Market Perform rating on Datadog and kept its price target at $237. Against a current share price of $283 cited in the report, that implies about 16% downside.
The note is clear that the rating is not a critique of the business itself. Bernstein says the issue is valuation and how crowded the long trade has become. In its view, some investors still expect growth to keep accelerating. If growth instead stabilizes or moves lower over the next few quarters, disappointment could show up first through less conviction in those crowded bullish positions.
By Bernstein’s figures, the stock trades at 105x 2026 adjusted earnings and 82x 2027 adjusted earnings, with an enterprise value to sales multiple of about 17.5x. For a company that may see revenue growth move from 35% to about 33%, the report argues those multiples leave little room for any softening in expectations.
Long-term drivers remain intact, but Bernstein says they are already priced in
Bernstein does not dispute Datadog’s long-term business case. The report says the combination of observability and security, AI-driven security analytics, and the appeal of its BYOC, or Bring Your Own Cloud, architecture for large customers all remain valid.
As described in the note, Datadog is using BYOC to serve customers with very large log volumes, high costs, or strict security requirements. The architecture lets those customers access analytics and control-plane functions without centralizing data inside Datadog’s SaaS environment. Management sees that as an important path to winning larger deployments.
The report also says training workloads are becoming a larger part of AI customer usage. In recent quarters, new AI-related use cases have included model training as well as inference and production applications.
Still, Bernstein’s conclusion is that the market already reflects those positives, and perhaps more than fully. Its bottom line is not that Datadog lacks quality. It is that the stock may have already discounted much of what investors expect to hear over the next 12 to 18 months.
Disclosure in the source article
The original article states that it was a summary and interpretation by Chaoxiang Research of a third-party broker report from Bernstein, dated Aug. 6, 2026, combined with public market information. It also says the ratings, target price, earnings forecasts, and related judgments cited in the piece were the views of Bernstein analysts and represented only the position of that institution, not the view of Chaoxiang Research, and did not constitute investment advice. The source also notes that market decisions should be made independently and that the article should not be used as a basis for buying or selling securities.