
With the rise of AI Agents, many e-commerce platforms have introduced the Model Context Protocol (MCP) framework, allowing AI to securely connect to merchant back-end data and tools, such as databases, web pages, documents, APIs, and business systems, breaking the limitations of chatbots at the retail level. The online store platform SHOPLINE was the first to open the MCP framework in June this year and announced the results of the first phase of applications yesterday, demonstrating how AI can transition from concept to daily operations.

Customer Preferences for Anthropic and OpenAI
As of actual usage data in August this year, as well as a first-phase survey of merchants, local merchants' top four demands for AI operational automation are "AI Deep Data Analysis" (75%), "AI Reading Store Data" (72%), "AI Smart Listing" (65%), and "AI Marketing Automation" (65%). In terms of tool preferences, nearly 40% (36.8%) of MCP calls came from the Anthropic ecosystem, followed by OpenAI/Codex (34.2%).
SHOPLINE's General Manager in Hong Kong, Ken Kan, pointed out that among the three major popular application scenarios, merchants are most concentrated on "Inventory and Order Operations" and "Member Segmentation and Precision Marketing", each accounting for 25%; "Data Analysis and Insights" accounts for 15.6%. The MCP framework underwent a significant upgrade in September this year, extending the automation capabilities of AI agents to multiple areas such as order operations, customer management, product operations, promotional marketing, recommendation activities, and message customer service handling. As of mid-November, the activation rate of invited merchants' MCP exceeded 30% (32.1%), with 66% becoming high-frequency core users.

Traditional Seafood Brand Explores Online Shopping
Ken Kan demonstrated in a press conference how to add the MCP server in ChatGPT, where the first step is to open the Settings page, then click on Integrations, select Plugins, press the Add button in the upper right corner, and input the MCP server URL. Next, log in to the account and complete two-factor authentication. After connecting, one can query back-end data, such as the latest orders for verification. At this point, merchants can analyze order status over the past 30 days, time distribution, and hot-selling categories. Finally, based on insightful data, a promotional draft can be quickly generated, which the store reviews and manually lists.
Established in 1928, the traditional seafood brand "Wanshun Cheung" emphasizes that innovation is essential to align with the new era and avoid being eliminated. The platform holds three live broadcasts every week and now, using MCP, significantly saves time while simultaneously obtaining data instantly to quickly validate market acceptance.
One-Person Store Reduces Listing Time by 80%
Regarding the ACG GO, a specialty store for anime peripherals operated by one person, founder Huang Weilun stated that after using AI technology and MCP, the listing time has been reduced from 3 hours to half an hour, saving over 80% of the time, allowing approximately 90 products to be processed in a day, enabling focus on high-value work.
In addition to establishing an "AI Academy" and launching an "AI Merchant Co-empowerment Program" to help merchants master MCP technology, SHOPLINE is trialing a standard operating procedure (SOP) manual specifically for AI agents, gathering AI prompts modules aimed at e-commerce scenarios, allowing merchants to import exclusive skills with one click, including post-live order aggregation, tracking unpaid customers, and reordering hot-selling products. Combining no-code tools "Widget Library" and "AI Listing Assistant," the product add-on rate increased by 56%, and the checkout conversion rate grew by 36%.
This article is reprinted from the Xinhua News Agency, with copyright belonging to the original author and original media.
免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到[email protected],本平台相关工作人员将会进行核查。