Will Datadog's AI-native customer cohort exceed 20% of total ARR in any quarter reported by H1 2026?

The denominator effect is the primary structural headwind. Non-AI revenue grew 23% YoY in Q4 2025 and the ex-largest customer business is guided at 'at least 20%' growth for FY2026. For the AI cohort to reach 20% of total ARR, AI ARR must grow substantially faster than the non-AI base — roughly 35-50% growth while total ARR grows 25%, starting from an estimated 12-15% share. The arithmetic requires AI ARR to grow from ~$450-570M to ~$840-900M in two quarters while total ARR grows from ~$3.8B to ~$4.3B. That is a ~75-100% AI ARR increase in one year while total ARR grows 14% — implausible given the OpenAI 'very conservative assumption' in FY guidance. The H1 window gives two bites (Q1 and Q2), but even Q2 2026 is likely to land at 15-17% rather than 20%.

I weight the disclosure risk alongside the fundamental trajectory risk. The company has not updated the AI cohort percentage metric explicitly in the Q4 2025 earnings call — the last precise disclosure was Q3 2025 at 12%. If Datadog continues to not explicitly report the AI cohort percentage in future earnings calls, resolution of this market may be ambiguous even if the underlying ARR trajectory improves. The sibling market lesson on voluntary disclosure (OpenAI concentration: 8% YES predicted, resolved NO) suggests management avoids explicit disclosure of sensitive concentration data. AI cohort percentage is directionally favorable (growth narrative), but management may choose not to frame it as a precise percentage. Even assuming full disclosure, the fundamental trajectory supports 15-18% by H1 2026 rather than 20%, requiring a significant acceleration above the ~1.2-2pp per quarter rate observed to date.

The OpenAI concentration issue is the most material single-variable risk. OpenAI is estimated at $170-240M ARR — approximately 60% of the total AI cohort. The CFO's 'very conservative assumption' on the largest customer in FY2026 guidance strongly implies management is modeling OpenAI ARR as flat or potentially declining. If the numerator's largest component is flat, AI cohort ARR growth depends entirely on the remaining ~40% of the cohort. The 19 customers at $1M+ ARR (from 15+ in Q3) and the new 8-figure financial model company deal add meaningful ARR but are insufficient to drive the cohort from ~13% to 20% of total ARR in two quarters when the denominator is simultaneously growing at 25%. I assign 13% probability — the direction is right but the magnitude required makes this a low-probability event by H1 2026.

I assign the highest probability among this ensemble because H1 2026 covers two quarters (Q1 and Q2 earnings), and Q2 2026 data would reflect the H2 deceleration in total revenue that the FY2026 guide implies. The guide shows Q1 at 25-26% growth but implies H2 deceleration — if total ARR growth decelerates to 18-20% by Q2, the denominator grows slower, making the 20% AI share threshold easier to reach. Additionally, the new 8-figure AI financial model company deal (the largest new logo ever) and continued AI customer count growth (+30% QoQ) add to the numerator. MCP tool calls grew 11x in Q4 vs Q3, and LLM observability spans grew 10x in 6 months — these usage intensity signals suggest AI customers are deepening spend. I still assign below 20% because even under an optimistic scenario, getting from ~13% to 20% in two quarters requires a step change, not a trend continuation.

The key analytical tension is that AI-native customer count grew 30% QoQ (500+ to 650) while AI ARR share appears to have stayed roughly flat or improved modestly (12% Q3 to an estimated 12-15% Q4). This divergence suggests new AI-native customers are being onboarded at below-average ARR — they are smaller, earlier-stage AI companies, not additional hyperscale spend. The 19 customers at $1M+ ARR growing from 15+ (27% growth) is more relevant to share movement than customer count growth, and even 27% growth at the $1M+ tier means adding roughly 4 customers at a combined ~$4-6M ARR — a rounding error on a $3.8B total ARR base. The step change needed to go from ~13% to 20% requires OpenAI or another hyperscale customer to dramatically accelerate spend, and the current guidance framing makes this the least likely scenario for H1 2026.

The NARRATIVE_REALITY_GAP = DISCONNECTED classification creates a structural paradox for this question: the committee's core finding is that the 'AI is the only driver' narrative is wrong and non-AI growth is robust. If non-AI remains robust at 23%+, the denominator prevents AI from reaching 20% of total ARR. The better Datadog's core business performs — which is the committee's central thesis — the harder it is for AI to cross the 20% threshold. I assign 17% accounting for two non-negligible upside scenarios: (1) OpenAI significantly exceeds the 'very conservative' assumption and adds $100M+ in new ARR; (2) the AI financial model company 8-figure deal triggers additional enterprise AI deals in Q1-Q2 2026 that aren't captured in current guidance. Both are possible but not the base case, consistent with a 15-18% probability.

Simple rate-of-change math: AI cohort grew from 8.5% to 12% in Q1-Q3 2025, approximately 3.5pp over 2-3 quarters, or roughly 1.2-1.8pp per quarter. To reach 20% from 13% (Q4 2025 midpoint estimate) requires 7pp of growth in 2 quarters at H1 2026. At 1.5pp per quarter, that is 3pp of growth — reaching 16%, not 20%. The rate would need to approximately triple to hit 20% in Q1 2026 or double to hit 20% by Q2 2026. No data in the context supports a tripling or doubling of the AI cohort share growth rate. Denominator growth alone (non-AI at 23% YoY) would prevent rate acceleration even if AI ARR grew at 30%. I assign 12% as the low-probability scenario where an extraordinary combination of OpenAI acceleration, new large AI deals, and non-AI deceleration simultaneously occurs.

The H1 deceleration scenario embedded in FY2026 guidance provides a small upside path. If Q1 2026 total revenue grows at 25-26% (as guided) but Q2 2026 decelerates to 18-20% due to the 'very conservative' largest customer assumption in H2, then by Q2 2026 the total ARR base grows more slowly. At the same time, AI customers continue to expand (MCP tool calls 11x, LLM observability 10x in 6 months) — AI ARR might grow 25-30% while total ARR grows 18-20% in Q2, allowing share to increase more rapidly. This scenario supports reaching 17-18% by Q2 2026, still below 20% but narrower than the base case. I assign 15% to account for this partial upside and the non-zero possibility that an additional hyperscale AI deal (similar to the 8-figure Q4 land) closes in Q1-Q2 2026.

Disclosure risk is a distinct and underappreciated factor. The company did not provide an explicit AI cohort percentage update in the Q4 2025 earnings call — marking the first reporting period without an explicit percentage disclosure since the metric was introduced in Q1 2025 (at 8.5%) and Q3 2025 (at 12%). If this disclosure becomes less frequent or is discontinued, the market may not be resolvable even if the underlying trajectory reaches 20%. The Bits AI SRE agent reaching 2,000+ trial customers in its first month suggests Datadog is pivoting its own AI narrative toward agent productivity rather than AI-native customer share — management may de-emphasize the AI cohort percentage metric in favor of newer KPIs. This non-fundamental resolution risk adds additional downward pressure beyond the trajectory math, supporting a 13% probability.