Marc Chaikin’s #1 AI Stock for 2027

Datadog reports Wednesday after the close, and the past week has made this a significantly harder room to walk into.

ServiceNow fell sharply after its Q3 results and outlook disappointed investors. Snowflake dropped more than 7% on October 30. Datadog itself fell 11.5% that same Friday, closing at $235.27, with no company-specific news behind the move. The drop was pure sentiment: investors growing uneasy about whether AI capital expenditure will ever produce the consumption growth these valuations require.

That question lands in Datadog’s lap Wednesday evening.

What the Numbers Need to Show

Analysts are forecasting Q3 revenue of $1.14 billion, up roughly 29% year over year, alongside EPS of $0.64. Those figures sit inside the guidance Datadog itself issued after Q2, when the company posted $1.12 billion in revenue and $0.65 EPS against a $0.58 consensus, then raised its full-year revenue outlook to $4.45 to $4.47 billion. The street is essentially asking Datadog to deliver what Datadog promised.

Several analysts expect it to exceed that bar. Oppenheimer has said it sees Q3 revenue coming in more than 4% above current consensus. Rosenblatt reaffirmed its Buy and $305 target on October 30, the same day the stock was sold alongside everything else.

But beating the number is only half the problem now.

Why Guidance Matters More Than the Beat

ServiceNow beat on revenue and still fell sharply. What broke the stock was forward commentary implying that demand was softening. Datadog’s forward earnings multiple sits well above 300, which means there is no cushion for ambiguity about what comes next.

Workday ran the same script one quarter earlier. Enterprise software investors treated an in-line forward guide as a miss, selling a stock that beat on every reported metric.

The bull case here is specific: AI workloads generate real, measurable, billable consumption on Datadog’s observability platform. The company has reported AI-native customers now represent 12% of total revenue, up from 6% a year earlier, with that cohort growing at roughly 253% year over year. LLM observability, GPU monitoring, and agentic workflow tracking are the products doing that work.

Wednesday’s report is the moment those claims get stress-tested against actual dollars. Analysts are forecasting remaining performance obligations of $3.55 billion, up from $2.79 billion a year ago. That figure is a more reliable signal of enterprise conviction than the headline EPS. A shortfall there, or guidance that implies deceleration, runs the same playbook ServiceNow just ran.

How to Read the Session

One useful data point from October 30: estimates did not move much during the selloff. What moved was the multiple the market was willing to pay. That is a separate problem from the business itself, but it matters for how you size a position into a binary event.

The multiple compression on October 30 reflects a broader repricing already visible in credit markets. Bond investors have been widening spreads on AI-related issuers, signaling skepticism about whether AI capex translates into durable returns.

Disciplined traders will be listening to the conference call rather than reacting to the headline. Management’s language around AI customer growth and Q4 revenue guidance will carry more information than whether EPS comes in at $0.64 or $0.67.

The Trader’s Lesson

High-multiple growth stocks rarely crack because the company missed by a few cents. They crack because guidance exposes that the demand the valuation was pricing has started to slow. ServiceNow was a sharp lesson in that. Watch what Datadog says about what comes next, not just what already happened.