Open in interactive viewer → charts, metric popovers & call review

KE Holdings: AI Becomes a Direct Variable in the Consumer-Centric Pivot

China’s top housing platform posts 75% non-GAAP net income growth as it shifts from scale to efficiency, deepening AI integration across services.
BEKE · Earnings Call · 2026-08-21

Introduction: A Profitable Pivot

KE Holdings (BEKE) reported Q2 2026 results that demonstrate a decisive strategic rehabilitation. Total GTV grew 6.3% year-over-year while revenue declined 5.7%, but non-GAAP net income surged 74.9% to RMB 3.185 billion. As CFO Tao Xu explained, "Contribution margins across all core business lines improved year-over-year and quarter-over-quarter, driving the group’s gross margin up 6.7 percentage points year-over-year to 28.6%." This margin expansion, paired with a 14.1% reduction in operating expenses, underpins a three-year high in non-GAAP net margin at 13%. The company is executing a deliberate pivot from scale-driven growth to operating efficiency and profitability, with CEO Stanley Peng describing a transformation that is “just the beginning.”

The Direct Variable: AI as a Production Factor

The most striking shift is Peng’s framing of AI not as an efficiency tool but as a direct variable in the company’s production function. He declared:

Now back to the very first question, whether AI is a direct variable. Because it changes who we serve, our judgments, our process, and our organization. This is a direct variable. That means we’re not simply installing AI into the company. We are regrowing the company with AI.

Stanley Peng, Management · 2026-08-21
This is a fundamental departure from the prior narrative. A year ago, Peng framed AI as a tool that “does not weaken service professionals” but rather “differentiates them, amplifies the base events and makes the upgrading of the service capabilities itself, one of the most important growth levers” (“AI does not weaken service professionals. It differentiates them, amplifies the base events and makes the upgrading of the service capabilities itself, one of the most important growth levers.” — Yongdong Peng, Management · 2026-03-16). Today, AI is being embedded into the core workflow, enabling a shift from managing averages to managing individual properties, clients, and agents. This is visible in new roles like the “client manager” who is not paid per transaction, ensuring objectivity, and in the division of labor across consulting, showing, contracting, and even deep data collection. The company is betting on deep service and a consumer centric model where every specialized skill is independently verified and compensated.

Operational Alpha Across Segments

The profit surge is not a one-off. In existing home services, GTV grew 8% year-over-year to RMB 629.9 billion, but more importantly, contribution margin rose to 46.1% (up 6.1 points). Peng attributed this to "higher unit productivity within our stable network and a better conversion of platform service value into revenue," rather than network expansion. Non-Lianjia platform service revenue grew 27.8%, outpacing GTV, as professional marketing and facilitation earn their keep. In new home, GTV was stable but contribution margin climbed 4.4 points to 28.8% through better project selection and cost optimization. Home renovation revenue declined 30.1% due to deliberate exits from inefficient channels, yet contribution margin jumped 7.5 points to 39.6% on centralized procurement. Meanwhile, rental services shifted to a lighter net-based model; managed units exceeded 790,000 (up 34%), with contribution margin up 6.9 points to 15.3%. Chief among the drivers is a structural shift toward net based product recognition and improved operational productivity. The market backdrop is nuanced. Existing home transactions are recovering, with tier-one city volumes outpacing others, but new home demand remains under pressure. Peng noted that “the market remains polarized” and that customers now require “professional judgment and transaction certainty.” This is precisely the value prop BEKE is scaling: using AI and data to deliver decision support beyond simple matching. The company’s forward stance is clear: it will not return to extensive expansion, but will allocate resources dynamically to protect margins and cash flow.

Market Context and Outlook

While the global tape shows AI-related keywords like “AI data centers” and “High performance computing” surging, BEKE is applying AI to a different frontier—real estate services. The company reported a strong second quarter but also acknowledged that this is "step one" of a longer transformation. The immediate results validate the strategy: non-GAAP operating margin hit 14.6%, up 8.5 points year-over-year. As Peng concluded, "Directional matters require unwavering bets"—and the bet on AI as a direct variable is now reflected in both operations and earnings. With a broad cash balance of RMB 67.3 billion and consistent share repurchases, BEKE has the balance sheet to fund this transition. The market will be watching whether the 74.9% net income growth is sustainable across quarters, but the evidence today suggests a company that has fundamentally changed its operating model—not just its cost structure. As Peng put it, "Q2 is not the conclusion, it is just the beginning."