To The Moon

On August 13, 2026, YXT.com( $云学堂(YXT)$ ) Group Holding Limited delivered a notable set of first-half results. Revenue reached RMB162 million, up 6.0% year over year. Gross margin rose from 65.1% a year earlier to 70.1%, an increase of 5 percentage points. Net loss narrowed sharply from RMB73.9 million to RMB14.4 million, representing an 80.5% year-over-year decrease.

For an enterprise services company in the middle of a strategic transformation, the message is clear: operating quality is improving, AI capabilities are beginning to translate into revenue, and the efficiency gains from AI are starting to show up in the numbers.

From Streamlining to Strengthening: Structural Improvement in Financial Quality

Over the past few years, YXT has been streamlining its business by shifting its customer mix toward large enterprises, scaling back non-core operations, and allocating more resources to subscription-based solutions and AI-related products. By the first half of 2026, the benefits of that restructuring were beginning to emerge.

The most direct evidence came from customer data. As of June 30, 2026, the number of subscription customers increased from 2,358 a year earlier to 2,391. More importantly, net revenue retention for subscription customers rose from 100.3% to 102.6%. An NRR above 100% indicates that existing customers are not only being retained, but are also expanding their spending. For a subscription-based business, this is an important indicator of customer stickiness and expansion potential. The nearly 50% year-over-year increase in newly signed customers also suggests that YXT’s AI-enabled products are gaining stronger traction in the market.

YXT’s revenue quality also appears to be improving. Within corporate learning solutions, subscription revenue reached RMB152 million, accounting for a larger share of the segment’s total revenue. This suggests that the Company’s revenue base is becoming more stable. Revenue from non-subscription offline training solutions declined from RMB7.7 million to RMB6.4 million. The contrast reflects a continued shift from project-based revenue toward subscription-based revenue, improving both predictability and sustainability.

Cost optimization was also notable. General and administrative expenses fell sharply from RMB54.2 million to RMB11.9 million, down 78.0% year over year. R&D expenses, however, increased 9.8% to RMB53.0 million. The combination of lower G&A expenses and higher R&D spending suggests that YXT is reducing its administrative cost base while continuing to invest in AI product capabilities.

Edison also raised its R&D cost assumptions in its latest research report, reflecting continued investment in AI product capabilities. Sales and marketing expenses declined from RMB61.9 million to RMB60.1 million, as AI tools helped improve efficiency across customer acquisition, conversion and retention.

Overall, YXT’s net loss narrowed from RMB73.9 million to RMB14.4 million, while adjusted net loss decreased from RMB64.0 million to RMB12.2 million, down 80.9% year over year. The scale of the improvement suggests that YXT’s loss reduction was not simply the result of cost-cutting, but rather a combination of revenue growth, gross margin expansion and operating efficiency gains.

Edison noted in its report that “the use of AI internally is helping to improve operational efficiency, resulting in a lower operating cost base.” The firm also forecasts that YXT will achieve positive EBITDA and operating profit in FY28. In other words, under Edison’s assumptions, YXT may be less than two fiscal years away from a potential profitability inflection point.

AI Commercialization Is Moving from Story to Numbers

Financial results show what the Company achieved over the past six months. But the commercialization of its AI business is likely to be a key measure through which the market assesses YXT’s long-term value.

In the first half of 2026, monthly recurring revenue from AI-related products reached RMB4.4 million, compared with RMB0.5 million in the same period last year, representing nearly nine times growth. SaleSmart, the Company’s AI-powered sales intelligence solution, generated more than RMB5 million in sales during the reporting period. While the absolute scale remains early-stage, the growth rate suggests that YXT’s AI products are moving from proof-of-concept toward measurable commercialization.

Customer cases also show that AI is producing measurable results in real business scenarios. A publicly listed home renovation services company in China improved its first-time customer store-visit conversion rate by more than 20% after adopting SaleSmart, while saving more than RMB100,000 annually. A rural commercial bank with more than 400 branches deployed an intelligent talent development system, shortening the average time for new employees to work independently from about one month to around three weeks. Customer complaints in the first month declined by 30%, while the probation pass rate improved by approximately 15%.

These cases suggest that AI adoption is moving beyond the question of whether a product has AI features. The more important question is whether AI can improve specific business outcomes.

Beyond front-end product commercialization, the AI-native transformation of YXT’s R&D system is another important, and potentially underappreciated, signal. In July 2026, YXT announced that it had completed an AI-native upgrade of its R&D system, enabling AI to play a deeper role in code development, maintenance and the broader R&D process, marking a new stage in its development as an AI Native Enterprise.

The Company has completed the construction of an ontology knowledge base for its core products, built on 15 years of product accumulation. This allows AI to better understand the historical architecture, design decisions and business logic behind more than 40 million lines of code. In practice, AI is no longer merely a coding assistant; it is becoming part of the R&D workflow, including requirements analysis, architecture design, code generation and testing validation. R&D teams can then focus more on specification definition, business judgment and key decisions.

The long-term significance of this underlying capability may be greater than the revenue growth of any single product. Improved R&D efficiency can translate into faster product iteration and a more efficient cost structure, while reusable knowledge assets can become a long-term competitive advantage. When the history and logic of a complex software system can be systematically understood and accessed by AI, future development and maintenance efficiency may change meaningfully.

Edison also provided a longer-term perspective. Based on a discounted cash flow analysis using an 11.5% weighted average cost of capital and a 2% long-term growth rate, Edison generated a base-case valuation of US$13.14 per ADS. Based on YXT’s pre-market share price on August 18, this implies upside potential of more than 270%.

Conclusion

YXT’s first-half financial performance may be only the beginning.

As its AI product portfolio continues to expand and commercialization accelerates, AI-related revenue could begin to show greater scale effects. At the same time, subscription customer growth and net revenue retention may continue to improve.

For a technology company undergoing a broad AI transformation, YXT’s profitability inflection point may arrive earlier than the market expects.

About YXT.com

YXT.com (NASDAQ: YXT) is a technology company focusing on enterprise productivity solutions. With a mission to "Empower people and organization development through technology," the Company strives to become the supreme provider in building and boosting enterprise productivity by combining over a decade of experience in tech-enabled talent learning and development and with AI-augmented task copilots and unleashing the power of knowledge and synergy. Since its inception, YXT.com has supported and received recognition from numerous Global and China Fortune 500 companies.