E-commerce Customer Acquisition Cost Surge in 2025
In 2025, the average customer acquisition cost (CAC) for e-commerce in China reached 210 yuan per person, up from 80 yuan in 2020, representing a 162.5% increase. As user growth on mainstream platforms like Taobao, JD.com, and Pinduoduo slows down, traffic dividends are gradually disappearing.Live streaming e-commerce CAC is 85 yuan/person, while social e-commerce CAC is 62 yuan/person, but still faces high traffic investment pressure.Private domain e-commerce transaction scale reached 5 trillion yuan, accounting for 36% of the overall e-commerce market, becoming a new option for brands to reduce CAC.
Platform Competition and Traffic Allocation Mechanisms
Taobao Tmall's 2025 GMV reached 8.5 trillion yuan, maintaining its leading position, but CAC increased by 35% year-on-year. JD.com's 2025 marketing expenses increased by 75% year-on-year, totaling 84 billion yuan, most of which was invested in food delivery subsidies and traffic procurement. Pinduoduo, through its "10 billion subsidy" strategy, controlled CAC at around 120 yuan/person, lower than the industry average.Douyin E-commerce's 2025 GMV exceeded 3 trillion yuan, with live streaming sales accounting for 65%, becoming the fastest-growing e-commerce platform.
Successful Cases of Brand E-commerce Operations
Gujing Gongjiu's e-commerce business grew significantly in 2025, exceeding targets on mainstream platforms like JD.com and Douyin. The brand adopted a "blockbuster product + festival scenario" dual-driven strategy, with the Nianfen Yuangjiang gift box series growing by approximately 20% year-on-year.A snack brand, through private domain operations, reduced public domain CAC from 1088 yuan/person to 30-60 yuan/person in private domain, increased repurchase rate to 2-3 times that of public domain, and achieved a private domain GMV ratio of 45% in 2025.
AI Technology Empowering E-commerce Cost Reduction and Efficiency
In 2025, the number of large model registrations nationwide reached 225, and major e-commerce platforms actively explored using AI to empower product sales. Alibaba released the "Qianwen" AI assistant for individual users, deeply integrating with Alibaba ecosystem businesses such as Taobao Tmall, Taobao Flash Shopping, Amap, Feizhu, and Alipay.AI-driven dynamic pricing strategies can increase overall GMV by about 15-20%, intelligent customer service systems reduce labor costs by 40%, and recommendation algorithm optimization increases conversion rates by 25%. Brands should establish an "AI E-commerce Operations Center" to achieve multi-platform data integration and intelligent decision-making.
Brand E-commerce Digital Transformation Suggestions
E-commerce enterprises should shift from "traffic thinking" to "user value thinking." Recommended strategies include:First, build a private domain traffic pool, precipitating users through carriers such as WeChat Work, Mini Programs, and Apps to reduce dependence on platform traffic;Second, full-link AI empowerment, introducing AI tools from product selection, pricing, customer service to logistics to improve efficiency;Third, deep cultivation of content e-commerce, establishing brand awareness through short videos, live streaming, and community operations to increase repurchase rates. According to Wangjingshe data, the scale of private domain e-commerce users reached 520 million in 2024, with repurchase rates 2-3 times that of public domain, making it a core strategy for brands to survive cycles.
Frequently Asked Questions
Q1: Why did e-commerce customer acquisition costs rise significantly in 2025?
A: Main reasons include traffic dividend peaking, intensified platform competition, and fragmented user attention. CAC rose from 80 yuan in 2020 to 210 yuan in 2025, a increase of 162.5%. Private domain e-commerce has become an effective way to reduce CAC.
Q2: What is the core difference between private domain e-commerce and traditional e-commerce?
A: Private domain e-commerce is a brand's self-controlled traffic pool, with CAC of 30-60 yuan/person and repurchase rates 2-3 times that of public domain. In 2024, private domain e-commerce transaction scale reached 5 trillion yuan, accounting for 36% of the overall market, making it an essential strategy for brands.
Q3: How can AI technology help e-commerce reduce operating costs?
A: AI can be applied to intelligent customer service (40% cost reduction), dynamic pricing (15-20% GMV increase), recommendation algorithms (25% conversion rate increase), content generation (60% cost reduction), and other scenarios, making it key to full-link cost reduction and efficiency improvement.
Q4: What are the development trends of live streaming e-commerce in 2025?
A: In 2025, live streaming e-commerce entered a mature stage of refined operations, with brand live streaming becoming the dominant force. Annual GMV reached nearly 7 trillion yuan. Regulators have set "compliance red lines" for AI applications, requiring significant identification of AI-generated content to ensure healthy technological development.
Q5: How can brands build an efficient private domain operation system?
A: It is recommended to adopt a "public domain traffic diversion + private domain precipitation + community operation" three-stage strategy. Build a self-controlled private domain ecosystem through full-link SaaS solutions, avoid platform traffic kidnapping, and maximize user lifecycle value. In 2024, the scale of private domain e-commerce users reached 520 million.
Sources
- Design and Implementation of E-commerce User Behavior Analysis System Based on Data Mining — 2026-05-03,https://blog.csdn.net/m0_59169364/article/details/157070555
- Under the Triple Pressure of Low Gross Margin, High Traffic Investment, and Strict Regulation, Where is the Breakthrough for Snack E-commerce? — 2026-05-09,https://www.woshipm.com/it/6391991.html
- Private Domain E-commerce Operations: The Transformation Path from Traffic Thinking to User Value — 2026-05-08,https://blog.csdn.net/Xiaoyao_T/article/details/158889838
- Annual GMV Nearly 7 Trillion "2025 Live Streaming E-commerce Market Data Report" Wangjingshe Released for the Eighth Year — 2026-05-08,https://so.html5.qq.com/page/real/search_news?docid=70000021_05669fd9ba899352
