Near-term expectations are for another acceleration. Jefferies expects the Cloud and Compute Services segment—combining Alibaba Cloud and chip unit T-Head—to grow more than 50% year on year in the September quarter, with momentum strengthening through the quarter ending March 2027. That is a forecast, not company guidance, and is especially sensitive to enterprise AI adoption and capacity deployment.

Wu’s payback framework implies an increasingly viable investment case. At current average gross margins, he expects the AI-computing investment to break even in about three years; if gross margins improve, the payback period could compress to about two years. Higher-margin AI products are therefore central both to the revenue outlook and to recovering the capital invested.

The trade-off is real in the short term. Alibaba has reached roughly the halfway point of its 380 billion yuan, three-year AI-infrastructure programme, while capex rose sharply and June-quarter profit fell 75%. The company had also previously indicated that investment could exceed the original 380 billion yuan target.

What this means for the longer term: if cloud growth stays above 50%, infrastructure utilisation rises, and AI-product margins remain above the cloud average, Alibaba could emerge with faster cloud revenue growth, expanding segment operating margins, and improving returns on invested capital after the heavy build-out. Conversely, weaker demand, excess capacity, or AI price competition would delay break-even and keep pressure on free cash flow.

Near-term expectations are for another acceleration. Jefferies expects the Cloud and Compute Services segment—combining Alibaba Cloud and chip unit T-Head—to grow more than 50% year on year in the September quarter, with momentum strengthening through the quarter ending March 2027. That is a forecast, not company guidance, and is especially sensitive to enterprise AI adoption and capacity deployment.

Wu’s payback framework implies an increasingly viable investment case. At current average gross margins, he expects the AI-computing investment to break even in about three years; if gross margins improve, the payback period could compress to about two years. Higher-margin AI products are therefore central both to the revenue outlook and to recovering the capital invested.

The trade-off is real in the short term. Alibaba has reached roughly the halfway point of its 380 billion yuan, three-year AI-infrastructure programme, while capex rose sharply and June-quarter profit fell 75%. The company had also previously indicated that investment could exceed the original 380 billion yuan target.

What this means for the longer term: if cloud growth stays above 50%, infrastructure utilisation rises, and AI-product margins remain above the cloud average, Alibaba could emerge with faster cloud revenue growth, expanding segment operating margins, and improving returns on invested capital after the heavy build-out. Conversely, weaker demand, excess capacity, or AI price competition would delay break-even and keep pressure on free cash flow.