Cloud revenue, meaning money corporations pay to rent computing infrastructure and software from a provider rather than own and run it themselves, crossed $100 billion at Microsoft's Azure division for the full fiscal year. Microsoft also beat fourth-quarter cloud expectations. The figure that deserves the most scrutiny sits in the backlog: signed customer contracts not yet converted to billed revenue grew 8% over the prior quarter, and that growth came mainly from businesses outside the large artificial intelligence sector.
What a backlog measures and why this one matters
A backlog is not revenue. It counts contracts already signed by customers but not yet delivered and billed, which makes it a leading indicator of future revenue rather than a record of what the company has already earned. When a backlog grows faster than revenue, it means new commitments are arriving faster than old ones are being worked off and billed.
An 8% quarter-over-quarter increase is worth noting on its own. The more telling detail is where it came from.
Why the customer mix changes the read
The backlog expansion came mainly from customers other than large AI companies. That one line reframes the headline.
Cloud spending in recent years has been closely associated with AI infrastructure demand. Large technology companies building and operating AI systems have been a meaningful source of cloud capacity purchases. Microsoft's backlog growth this quarter came from elsewhere.
A backlog driven by a narrow group of large buyers carries concentration risk: if those buyers reduce spending, the contracted pipeline shrinks quickly behind them. Growth spread across a wider customer base is harder to unwind. Microsoft's 8% expansion came from that broader base.
What was signed versus what is only projected
The $100 billion Azure number is recognized revenue, money billed and collected across the full fiscal year. The backlog is a different category: binding customer commitments that have not yet been converted into reported results.
Microsoft beat fourth-quarter cloud expectations. The backlog grew 8% over the prior quarter. The customers driving most of that growth were outside the large AI sector.