Datadog grew revenue 35% in Q3, but its Q4 outlook implies virtually no sequential expansion. The unnamed AI customer's consumption reduction tests how much a large software contract protects against volatile usage.
Datadog's largest customer, an unnamed leading AI company, reduced software consumption during the third quarter of 2026. Datadog nevertheless reported $1.19709 billion of revenue, up 35% from a year earlier. The Q4 outlook is weaker: fourth-quarter guidance of $1.19 billion to $1.20 billion implies almost no sequential growth.
Q4 guidance points to almost no quarterly growth
Datadog reported $2.12788 billion in first-half revenue and $1.19709 billion in Q3, a combined $3.32497 billion through September. Subtracting that amount from full-year revenue guidance of $4.515 billion to $4.525 billion yields $1.19003 billion to $1.20003 billion for Q4, independently reproducing the quarterly guide after rounding.
At the $1.195 billion midpoint, Q4 would be $2.09 million, or about 0.2%, below Q3, on CTOL's arithmetic. Guided year-over-year growth of 25% to 26% is below the third quarter's 35%. This is a reconciliation of consolidated guidance, not a measure of revenue lost from the largest customer.
In its Q2 2026 Form 10-Q, Datadog warned that reduced usage by its largest customer beginning in Q3 could decelerate growth. On the Q3 earnings call, CFO David Obstler said Q3 consumption remained substantially above the contractual minimum. Management's Q4 forecast assumes the customer's revenue contribution at that minimum, although the company has not said Q4 usage will actually fall that far.
Management also built in an expected December usage slowdown and the effects of early renewals by other major AI customers that raised contractual commitments while giving those customers better unit prices. Obstler said growth excluding the largest account accelerated and revenue growth among non-AI customers reached the high-20% range. Without dollar weights for those effects, the Q4 guide cannot be apportioned to any one of them.
What the nine-figure renewal tells us
Earlier management described this customer's renewal as a nine-figure contract and said the account used 17 Datadog products. Those are meaningful commercial ties, but the nine-figure sum is not identified as annual revenue: Datadog has not disclosed the agreement's duration, precise minimum, discounts or overage spending. Platform breadth can coexist with lower metered usage.
Datadog's accounting disclosures describe commitments recognized ratably, usage commitments recognized as services are delivered, monthly usage subscriptions, and charges for consumption above commitments. Revenue depends on metered units, realized unit prices and the amount covered by contractual minimums.
Reducing metered activity need not reduce recognized revenue dollar for dollar, or in the same quarter, and the disclosed information does not provide the inputs for an account-level calculation.
Q3 free cash flow was $251 million, 21% of revenue, compared with a 26% trailing-12-month free-cash-flow margin. Those figures show consolidated cash production, but collections and payments vary by quarter. The approximately 80% Q3 GAAP gross margin was unchanged from a year earlier, but it cannot determine whether the customer's reduction helped or hurt profit after serving costs.
Customer counts cannot show revenue concentration
Datadog reported 33,500 total customers, 4,860 with at least $100,000 in annual recurring revenue, more than 850 AI customers and 39 AI customers with more than $1 million in annual recurring revenue. The 39 are a subset of the AI customer group; neither count reveals how sales are distributed among customers.
More revealing is the Q3 filing's statement that the AI-native cohort, including the largest account, contributed high-single-digit percentage points of consolidated year-over-year revenue growth. That establishes the cohort's importance, but not the largest customer's share. Trailing net revenue retention in the low 120s and remaining performance obligations of about $3.85 billion, up 38%, also mix customers and timing; they cannot reveal a recent account-level pullback.
Earlier consumption slowdowns offer the closest parallels
Datadog's own Q3 2023 commentary described cloud-cost optimization holding back consumption growth even as revenue rose. MongoDB reported slower-than-expected Atlas consumption growth in May 2024 despite 32% Atlas revenue growth. These are examples of the consumption-to-revenue mechanism, not evidence of what happened inside the unnamed AI customer's workloads.
Gartner analysts Andrea Curry and Padraig Byrne wrote: “Observability costs continue to grow as telemetry volumes, operational complexity, and AI usage increase.” This provides a reason to scrutinize sampling, retention and unit prices as workloads grow, but Datadog has not attributed the account's reduction to any such choice.
One longer-term possibility is a split between inexpensive telemetry storage and higher-value investigation workflows. In June 2026, Datadog and ClickHouse announced an integration designed to pair ClickHouse storage with Datadog's log exploration. That shows a possible architecture, not evidence that this customer changed suppliers.
What could happen next
The most likely near-term outcome is Q4 revenue near the guided range while the business outside the largest account continues expanding. A plausible upside is usage above the minimum assumed in guidance or stronger spending elsewhere; the downside is that more AI customers optimize consumption or negotiate lower unit prices. Those are conditions, not modeled probabilities.
Over the following year, the question is whether other AI and enterprise accounts can absorb top-customer volatility without weaker margins or retention. Over three years, the issue is whether growing observability workloads translate into higher realized revenue or shift more spending toward lower-priced telemetry storage. Neither path can be forecast from customer counts alone.
How to test the forecast
Datadog's financial results for the quarter ended December 31, 2026, and its related annual filing, expected in early 2027, will test whether growth decelerates while trailing retention remains stable. Even a successful test would not identify the customer's spending or establish a wider AI demand reversal. With medium confidence, CTOL expects Q4 year-over-year revenue growth of no more than 28% and trailing net revenue retention of at least 120%; growth above 28% or retention below 120% would falsify the call.