Chart of the Day | Coatue Summary

The headline claim frames the entire chart: "GenAI spend via cloud providers has surged 15–20x across every cohort in just 18 months!" This isn't isolated to top spenders — the chart shows growth across the Top 1%, Top 10%, and Median segments, suggesting broad-based adoption rather than concentration among a few whales.

Across all three panels (Top 1% Average, Top 10% Average, Median), spend was relatively flat/slow-growing from Jan '25 through mid/late '25, then bent sharply upward starting around late 2025 into 2026. The Top 1% cohort went from roughly $25-50K/day in early 2025 to over $300K/day by mid-2026 — an inflection pattern indicating the market shifted from experimentation to scaled production usage.

The Median panel shows growth from near $0K to over $1K/day per customer — a smaller absolute number, but proportionally in line with the top cohorts' 15-20x surge. This signals GenAI spend is diffusing beyond the largest enterprises into the broader customer base, not just a top-heavy phenomenon.

The data implies AI infrastructure demand may still be under-hyped rather than over-hyped, contrary to concerns about an AI spending bubble. The consistent 15-20x multiple across every cohort (not just outliers) — "a stable 07/212026 cohort-joined customer panel" — suggests durable, broad-based demand growth rather than a speculative spike concentrated in a few large deals.

Amazon Web Services (AWS), Cloud infrastructure provider, Mentioned as one of the three clouds where GenAI spend was captured, "Total AI spend includes Anthropic, OpenAI, Gemini, and open-source model spend captured across AWS, Azure, and GCP."
Microsoft Azure, Cloud infrastructure provider, Same as above — source of underlying spend data.
Google Cloud Platform (GCP), Cloud infrastructure provider, Same as above — source of underlying spend data.
Anthropic, AI model provider, Its model spend is included in the aggregated GenAI spend metric tracked in the chart.
OpenAI, AI model provider, Same — included in the aggregated spend metric.
Google Gemini, AI model provider, Same — included in the aggregated spend metric.
Yipit (Yipit Cloud), Data/analytics provider, Cited as the underlying data source for the chart: "Source: Yipit Cloud as of 07/212026. Coatue analysis as of August 2026."

None specifically named in the article (author attributed only to "Coatue Management" as an organization).

Track cohort-level spend trends, not just aggregate industry numbers. Coatue's methodology — segmenting into Top 1%, Top 10%, and Median — reveals that growth is broad-based rather than driven only by a handful of megacustomers, a more useful signal for entrepreneurs sizing market opportunity: "Daily spend calculated by dividing monthly spend per customer by the number of calendar days in each month... Data reflect a stable 07/212026 cohort-joined customer panel."
Watch for inflection points, not steady growth, when timing AI infrastructure or tooling bets. The chart shows a distinct hockey-stick starting in late 2025/early 2026 across all cohorts — operators building AI infra, tooling, or cost-optimization products should recognize that budget expansion is nonlinear and can happen suddenly once it starts.

The methodology note that "Median excludes customers with zero AI spend" is easy to skip but important: it means even the "median" figure likely understates how many total cloud customers still have negligible or no GenAI spend, implying the addressable expansion runway is larger than the chart alone suggests — full market penetration is still incomplete even as usage intensity skyrockets among adopters.