ChatGPT’s New Feature Aims to Take Over Taobao, Xiaohongshu, and Bilibili Tasks
The Overwhelmed Shopper Problem
Ever wanted to buy something, only to end up mentally exhausted after hours of research?
You read 20–30 posts on Rednote (Xiaohongshu), watch a dozen Bilibili reviews, check “What’s Worth Buying” discount channels… and finally give up, buying nothing.
---
A New Unified Entry Point from OpenAI
OpenAI has now introduced a Shopping Research feature in ChatGPT — a single place to handle all that “shopping homework” for you.
How to access it:
- In ChatGPT, click the menu `( + )`.
- Select Shopping research.
- Tell ChatGPT what you want to buy.
ChatGPT will:
- Search the web for product information.
- Compare specifications.
- Ask for your budget and preferences.
- Deliver a personalized buyer’s guide.
> Unlike standard search engines or price-comparison tools, ChatGPT uses your needs plus your chat history to find the single, best fit.
Availability:
As part of Black Friday and other holiday promotions, Shopping Research is now open to all logged-in ChatGPT users — Free, Plus, Team, and Pro. Pro’s Pulse feature also uses Shopping Research to offer tailored product choices.
---
Not Just Search — It’s Doing the Homework
Domestic platforms like Taobao/Tmall and JD.com already have AI shopping features based on your purchase history.
▲ Taobao AI Shopping screenshot
But inside a general-purpose chatbot, ChatGPT is the first to offer integrated shopping research. Past AI suggestions were hampered by broken or imaginary links — Shopping Research fixes that and boosts recommendation quality.
How It Works
- Enter Shopping Research mode.
- Example prompts:
- “Find me a quiet, wireless vacuum cleaner suitable for a small apartment.”
- Upload a dress photo: “Find similar styles.”
- ChatGPT behaves like a real sales assistant, asking follow-up questions:
- Budget range.
- Intended use.
- Key features valued.
- Gift or personal purchase.
- With Memory enabled, past preferences are remembered.
- If “gaming” is in your profile, laptop suggestions will prioritize graphics performance.
- After initial selections, you see visual product cards:
- Images
- Prices
- Specs
Interactivity:
- Swipe right or click More Like This for liked items.
- Swipe left or mark Not Interested for disliked items.
- Feedback shapes ongoing recommendations.
---
Personalized, Fast, and Context-Rich
During final report generation, ChatGPT gives educational product tips to make waiting painless.
Example for cameras: “Megapixels aren’t the only determinant of photo quality.”
The outcome is a full Buyer’s Guide:
- Popular product recommendations.
- Pros/cons.
- Key differences.
- Latest info from trusted retailers.
This approach replaces hours of manual research with minutes of guided conversation.
---
Extra Capabilities
- Image-based search: Find similar styles or models.
- Discount hunting: Includes student-exclusive discounts or subsidies.
---
Ideal Use Cases
OpenAI suggests high-decision-cost categories benefit most:
- Electronics: Phones, computers, cameras.
- Home & Garden: Vacuum cleaners, robot sweepers, furniture.
- Beauty & Skincare: Match ingredients to skin types.
- Sports & Outdoors: Camping gear, pro sports equipment.
Note:
- OpenAI does not take payments or secretly rank products based on sponsorships.
- Data comes from public web sources.
- Chat history is shared with no retailer.
---
Under the Hood
- Powered by a GPT‑5 mini model trained via reinforcement learning, specialized for shopping.
- New evaluation metric for product–user fit.
- Top recommendation accuracy: 64%.
- Prices and stock change quickly — verify via merchant link before purchase.
Future plan:
Direct purchases through ChatGPT using Instant Checkout (Stripe partnership).
Limitation:
Domestic Chinese brand data is sparse; real-time scraping isn’t possible yet.
International brand research works well, but Taobao/JD/Pinduoduo integration could come quickly.
---
User Reactions
Responses on X vary:
- “OpenAI implemented my entire startup idea.”
- “Can AI find me the best AI subscription?”
- “Stop ruining my ChatGPT.”
---
Bigger Picture — AI for Commerce
Recent ChatGPT updates (group chat, instant checkout, shopping integration, browser tools, adult mode) show OpenAI focuses on retaining its current massive user base.
Shopping Research is just a foothold in the e-commerce market — a small slice of a huge pie.
---
For Digital Creators & Entrepreneurs
Platforms like AiToEarn官网 demonstrate how shopping research is part of a larger AI commerce ecosystem:
Features:
- AI content generation.
- Multi-platform publishing.
- Analytics.
- AI model ranking.
Supported platforms:
Douyin, Kwai, WeChat, Bilibili, Rednote, Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, X (Twitter).
Outcome:
Turn AI research or creation into multi-channel impact and revenue — efficiently, at scale.
---
In summary: ChatGPT’s Shopping Research shifts AI from mere information source to interactive decision assistant. For complex purchases, it can condense hours of effort into minutes — while keeping control and transparency squarely with the user.
