Not a data repeater, but a microscope for truth. Peeling away the 'earnings report filter' of tech giants to dissect their real money-making secrets and growth anxieties.

Zero State LT proudly presents its earnings report analysis column, *Earnings Report Insight*, focusing on in-depth analysis of financial reports from the world's top tech companies. By stripping away noise and anchoring value, we explore long-term trends and business fundamentals, capturing decisive signals in the ocean of information to help you understand the next era.

This article examines Alibaba's Q1 FY2027 earnings report.

Author | Zhang Qian

Editor | Hu Zhanjia

Operations | Chen Jiahui

Produced by | Zero State LT (ID: LingTai_LT)

Header Image | Shot by Zero State LT

A 76% net profit decline in earnings reports may not necessarily be bad news.

This was my first impression after reviewing Alibaba Group's Q1 FY2027 earnings report released on the evening of August 20. The true focus is not the profit decline but Alibaba's complete overhaul of its reporting structure starting this quarter—dividing the company into four segments: E-commerce Group, AI Cloud & Computing Power, AI Labs & Applications, and 'All Others.' The change in ledger structure indicates the company wants the market to read it differently.

Let's first lay out the complete framework of this earnings report. For the quarter ending June 30, Alibaba reported revenue of RMB 268.953 billion, up 9% YoY; Net profit attributable to shareholders (net profit attributable to parents) of RMB 10.537 billion, down 76% YoY; and operating profit of RMB 15.161 billion, down 57% YoY. After the report's release, Alibaba's stock fell over 4% pre-market in the U.S., yet the next day, Hong Kong shares closed up 1.61%—the same numbers elicited two different market reactions.

The divergence between the two markets points to the same factor: Alibaba's new segment reporting structure.

The Profitable AI Cloud Enters Harvest Phase

The first segment to break through under the new structure is 'AI Cloud & Computing Power Services,' the new division formed after merging the Cloud Intelligence Group with T-Head Semiconductor.

According to the earnings report and comprehensive media information from Tianyancha, this quarter, the segment reported revenue of RMB 48.437 billion, up 45% YoY—a 22-quarter high—with AI-related product revenue reaching RMB 12.376 billion, marking the twelfth consecutive quarter of triple-digit growth. Based on earnings report data, annualized AI revenue has surpassed RMB 49.5 billion. More impressively, on the profit side: segment-adjusted EBITA soared 133% YoY to RMB 5.628 billion, with the profit margin rising from 7.2% to 11.6%. The simultaneous improvement in growth and profit margins indicates economies of scale are kicking in.

Two often-overlooked details underpin this segment's momentum. First, chips: T-Head has established a full-stack self-developed system covering GPUs, CPUs, storage, and network chips. Its latest-generation AI processor, Zhenwu M890, serves over 650 external clients across 20+ industries via Alibaba Cloud—a path validated by Amazon's self-developed chip business, which exceeds $25 billion in ARR, proving self-developed chips improve cloud profit margins, a strategy Alibaba is now replicating. Second, market share: Omdia's *China AI Cloud Market Share 2025* report shows Alibaba Cloud ranks first in China's AI cloud market with a 38.1% share.

In summary: Alibaba has started making money from upstream computing power sales, and the pace is accelerating.

The Costly Losses of AI Qianwen

Shifting focus to the next table, the tone changes completely.

The newly established AI Labs & Applications division—housing the AI Model Lab, Qianwen App, and Qianwen Office—reported revenue of RMB 3.338 billion this quarter, up 16% YoY, but adjusted EBITA losses ballooned to RMB 13.861 billion. Last year's loss was RMB 3.224 billion, a 330% increase in one year, with losses now 4.2 times revenue. Alibaba's explanation is straightforward: increased AI capability investments, coupled with rising inference costs for the Qianwen App.

The logic is clear: 250 million users have experienced AI shopping via Qianwen App, with each additional user call burning more inference computing power. The Qwen series models have surpassed 3 billion global downloads, with over 300,000 derivative models. In August, Alibaba open-sourced Qwen3.8-Max with 2.4 trillion parameters, trading open-source for ecosystem growth, which funnels API calls back to Alibaba Cloud—losses at the application layer, gains at the computing layer. That's Alibaba's calculation.

But whether this calculation pays off depends on maintaining model competitiveness and achieving paid conversions for Qianwen Office and Qianwen App. This earnings report offers no answers yet.

Where Did the Profits Go? Let's Break It Down

Instead of vaguely blaming AI for dragging down profits, let's dissect the -76% decline. The money primarily went to four places.

First, AI investments themselves: the AI Labs & Applications division incurred over RMB 10 billion in additional quarterly losses. Second, the 'All Others' segment swung from profit to loss—Alibaba Health, Digital Media & Entertainment, and Gaode's adjusted EBITA turned from a RMB 687 million profit to a RMB 3.343 billion loss, a RMB 4 billion swing. Third, one-time items: goodwill impairment of RMB 4.458 billion (zero last year) and approximately RMB 4.3 billion in EU fine provisions pushed general and administrative expenses from 3.0% to 4.7%. Fourth, hidden in the income statement: net interest and investment income halved from RMB 17.376 billion to RMB 9.004 billion, while income tax surged 44% to RMB 12.798 billion, with the effective tax rate jumping from 17.6% to 57.4%—last year's profits included substantial low-tax investment gains, which have now receded, altering the tax structure.

The divergence in the cash flow statement is equally noteworthy. Net cash from operating activities reached RMB 22.945 billion, up 11% YoY, indicating the core business remains resilient. However, free cash flow outflow widened to RMB 44.670 billion, more than double last year's deficit, with capital expenditures hitting RMB 67.678 billion, up 75% YoY—a quarter of quarterly revenue invested in AI infrastructure.

This profit-for-computing-power trade-off isn't unique to Alibaba.

Just last week, Tencent disclosed a nearly identical earnings report expression: Q2 capital expenditures of RMB 52.8 billion, up 176% YoY, with free cash flow turning negative by RMB 13.8 billion. China's tech giants have shifted from model competition to asset-heavy AI infrastructure races.

Conclusion Awaits Next Quarter

Laying out the profits of the four segments, Alibaba's capital cycle becomes clear: the E-commerce Group generated adjusted EBITA of RMB 39.749 billion, the sole large profit pool, funding both ends of AI—RMB 5.628 billion earned by Cloud & Computing Power with rising profit margins, and RMB 13.861 billion burned by AI Labs & Applications, with 'All Others' losing another RMB 3.343 billion.

Three months ago, Eddie Wu (Wu Yongming) stated that Alibaba's AI had moved past the initial investment phase into a positive scale-driven commercialization return cycle. This earnings report partially validates that claim—cloud growth and profit margins have indeed entered the return phase, but the application layer remains deep in the investment phase.

Thus, any conclusions now are premature. Three metrics warrant close attention: whether the share of AI-related product revenue in cloud external revenue can continue rising, whether cloud segment EBITA margins can stabilize above 12%, and when AI Labs' losses will begin narrowing. If these three factors reach inflection points, Alibaba's full-stack AI narrative will truly close the loop. If not, the -76% decline may not be growing pains but a signal for reevaluation.

This is the significance of the earnings report structure overhaul: Alibaba has extracted AI's financials from the ambiguous 'All Others' category, listing them separately with self-contained profit and loss.

Daring to clarify the ledger is, in itself, a statement.

For inquiries about this article, please contact

Follow Zero State's video channel to stay closer to business!