AWS Backlog Reaches $578B, Outpacing Amazon's Revenue Growth

The AWS backlog has climbed to $578 billion, expanding at a triple-digit pace. Its revenue, by comparison, rose 39.3% to $46 billion.

The company came in below earnings forecasts, and its stock slid more than 7% in extended trading. Free cash flow on a trailing basis swung to a $42.4 billion outflow from a $14.8 billion inflow a year earlier.

What explains AWS backlog expanding at a faster clip than revenue?

During the earnings call, management said the backlog increased by $82 billion versus the second quarter. Quarterly AWS revenue, on the other hand, grew by $3.8 billion.

Twelve months ago, that backlog was $200 billion. It has close to tripled even as AWS revenue advanced 39.3%. The sequential sales increase in the second quarter was $4.6 billion, making the most recent rise smaller.

Cloud firms are unable to convert order backlogs into revenue at the pace investors are looking for, according to Jefferies' Brent Thill.

He attributes the lag to bottlenecks in data centers and supply chains.

Capacity for 2026 demand was lacking, Amazon CEO Andy Jassy said in the prior quarter, and he expected the same situation in 2027.

Jassy added that an AI accelerator is being installed roughly every 14 seconds, and customers are taking up that capacity nearly as quickly.

Does 40% growth justify a heftier spending bill?

The company is now guiding to AWS growth of more than 40% in both 2027 and 2028. Growth has picked up speed for six consecutive quarters, after sitting at 20% in the third quarter of 2025.

For next year, Amazon plans to spend considerably more than the roughly $220 billion it set aside for 2026. That 2026 figure was raised from $200 billion in the previous quarter, with higher memory costs cited as the reason.

The third quarter alone accounted for $75.9 billion in cash capital spending.

Operating income at AWS, however, came in at $17.8 billion, up from $11.4 billion in the same period last year.

Microsoft and Alphabet had reported the day before, with AI infrastructure spending the central theme across all three big tech earnings reports.

Should delivery continue to lag orders, investors may size up hyperscalers by data centers switched on rather than contracts signed. Adding capacity faster would close that gap and take some pressure off the spending plan.