Google Cloud Customers Spend 50% More Than They Promised

Thomas Kurian stood before analysts and investors on July 24 and offered a number that explains, in a single figure, why Google’s cloud business is growing faster than its rivals’. Customers who sign cloud commitments with Google end up spending more than they promised, about 50 percent more on average, Kurian said.

The figure comes from the mechanics of the cloud business itself. Hyperscale providers sell capacity through committed-use contracts, in which customers pledge a minimum spend in exchange for discounts. When usage runs ahead of the commitment, the customer pays overage rates on top of the base contract. Google, Kurian said, has been collecting those overages at an unusually high rate.

“Our existing customers, once they commit, are spending well beyond what they committed,” Kurian said, describing the pattern across the company’s customer base. He credited the strength to a differentiated product portfolio and to execution in taking products to market, and said the trend is showing up in both revenue and operating profit.

The numbers back him up. Google Cloud’s revenue rose 82 percent from a year earlier, an acceleration that stands out even in a market where every major provider is posting double-digit growth on the back of artificial-intelligence demand. The unit, which spent years in losses after Google began breaking it out in 2018, reached operating profitability in 2023 and has been widening the margin since.

The growth is being driven by two forces. The first is AI infrastructure: enterprises are renting computing capacity to train and run models, and Google’s portfolio, including its tensor processing units, which are available to customers at a discount relative to rival chips, has become a default choice for price-sensitive AI workloads. The second is a broader migration wave, as companies move legacy applications off their own servers and into the cloud.

Kurian’s own history helps explain the trajectory. He joined Google Cloud in 2019 from Oracle, where he had spent more than two decades building the database giant’s product line, and took over a cloud business that was a distant third behind Amazon Web Services and Microsoft Azure, struggling to win enterprise trust. Under his tenure, Google Cloud rebuilt its sales force, standardized its product portfolio, and began winning the large enterprise and public-sector contracts that had long eluded it.

The competitive picture has shifted accordingly. AWS remains the largest cloud provider by revenue, and Azure is second, but Google Cloud has been gaining share in a market where the top three players collectively absorb the overwhelming majority of enterprise cloud spending. Analysts who track the industry said Google’s 82 percent growth rate reflects both a smaller base and real demand, not just easy comparisons.

The overage figure is worth dwelling on because it captures the difference between a growing business and a profitable one. Cloud economics reward utilization: the marginal cost of serving an additional hour of compute is low, so overages on committed contracts flow to the bottom line at high margins. A customer spending 50 percent more than committed is, in effect, funding margin expansion for the provider.

The spending pattern is not uniform across customers. Kurian said the overage trend is concentrated among the largest accounts, enterprises and AI startups that run continuous workloads, rather than smaller customers that buy capacity in bursts. That concentration is a risk as well as a strength: if a handful of big customers throttle back, the growth rate would normalize quickly.

The question analysts are asking is how long the surge can last. AI demand has been the industry’s dominant theme for two years running, and every hyperscaler has warned investors that the buildout of data centers is a bet with uncertain timing. Google’s advantage, Kurian argued, is that its customers are spending real money on workloads that are already in production, rather than on speculative capacity.

There are constraints. Google’s parent company, Alphabet, is spending tens of billions of dollars a year on data centers, and the capital intensity of the AI buildout has weighed on free cash flow even as the cloud business grows. Power availability, not demand, is becoming the binding constraint on capacity in many regions, and Google has signed deals for nuclear and geothermal energy to secure supply.

Kurian’s comments also signal a shift in how Google talks about its cloud business. For years, the unit was presented as an investment story, with growth emphasized over profit. Now the message is that both are moving together, and that the operating margin trajectory is the metric the company wants investors to watch.

The 50 percent figure, if it holds across the portfolio, has implications for the whole industry. Committed-use spending is the industry’s forward indicator, the number providers use to plan capacity years in advance. If Google’s customers are consistently over-spending their commitments, the signal is that demand is stronger than the contracts suggest, and that the buildout, expensive as it is, may be underpriced.

Related Posts

  • September 6, 2026
  • 10 views
Anthropic Moves Its IPO Filing to Late September

The bankers and lawyers running Anthropic’s initial public offering had told investors to expect the company’s registration documents as soon as this week. The calendar has moved. Anthropic now plans…

  • September 6, 2026
  • 9 views
Seattle Times and Newsday Sue OpenAI and Microsoft

The complaint filed Friday carries the tone of an elegy with a legal caption. The Seattle Times and Newsday, the Long Island daily, accuse OpenAI and Microsoft of scraping their…