ChatGPT Adds Virtual Try-On and Saved Products: An APAC Commerce Readiness Guide

What changed in ChatGPT shopping?

The “Try on” button appears on eligible clothing and accessory listings. A reference photo can be reused for later try-ons and managed under Personalization settings. The feature is available on mobile and web. Shoppers can also upload an item image in conversation and ask how it might look on them.

A bookmark saves a product to Favorites in ChatGPT Library, and web users can place saved products in folders. This lengthens the purchase journey beyond one answer. A user may return after a price, size or stock position has changed, which makes timely updates and clear effective dates more valuable.

Why the merchant page still matters

OpenAI says generated try-on images may not represent the product or shopper exactly and do not guarantee fit or size. Shoppers are directed to merchant measurements, product details and return policies before purchase. The visual experience may start inside ChatGPT, but the merchant’s evidence still carries the purchase conditions.

Brands should publish front, back and detail views, tie each image to the correct colour variant, and explain material, lining, stretch and measurement method in visible copy. Different names for the same product across a brand site and marketplaces can make comparison harder. Local teams also need to keep translations precise: a colour or fit label that works in one market may describe a different expectation in another.

How product results differ from ads

OpenAI states that ChatGPT product results are selected independently and are not ads or influenced by OpenAI partnerships. Ads are separate from product results. Some eligible products and merchants may show Instant Checkout, allowing a transaction to be completed without leaving ChatGPT.

Labels such as “Budget-friendly” or “Most popular” are generated from information available to the model and are not verified guarantees. Review summaries may draw on public websites, but OpenAI does not verify the underlying reviews and ratings. Brands cannot assume direct control over these descriptions; they can improve the public evidence by maintaining accurate specifications and responding to recurring customer issues.

What APAC teams should align

Pricing can come from third-party providers, and the first displayed price may not be the lowest. The same product may appear from several merchants at different prices. Regional teams should align currency, tax treatment, promotion dates, delivery fees, stock status and return windows for each market.

Shared identifiers help local data connect to the same product. Keep the brand name, product name, model number, GTIN or seller SKU stable across direct stores and major marketplaces. If Product structured data or merchant feeds are used, compare them with visible page content instead of treating feeds as a separate record. Korea also requires checks across NAVER and local marketplaces; Japan, Taiwan, Hong Kong, India and Australia each have different platform mixes, languages and fulfilment expectations.

Images, privacy and customer expectations

Virtual try-on uses a shopper’s reference photo. A merchant should not claim access to that photo without evidence, but any comparable feature hosted on the merchant’s own site needs a clear notice covering the controller, retention and deletion. Limitations about fit and representation should sit near the experience rather than being buried in a general policy.

If generated imagery is used on a product page, distinguish it from actual product photography. Differences in colour, length, trim or texture can increase returns and support contacts. Keep verified product images as the primary reference, label generated scenes as illustrative, and provide a direct path to measurements and return conditions.

A practical test set for AI shopping

Test queries should include a use case and constraint, such as “navy jacket for client meetings”, “running shoes for wet weather” or “a gift that ships quickly to Seoul”. Record whether the response identifies the current product, seller, price, availability and delivery area. A mention count alone cannot reveal whether the shopping answer is usable.

For eligible apparel, change colours and sizes and inspect whether the correct image and merchant link persist. Reopen saved products after several days to see whether price and stock have updated. When a discrepancy appears, compare visible copy, merchant feeds, structured data and marketplace records before assigning the cause.

AI Search Lab view

ChatGPT shopping now supports discovery, comparison, saving and virtual try-on within one continuing experience. The operational priority is consistent product evidence across every place an assistant may read. APAC headquarters can share data rules and tests, while local teams own language, price, availability, fulfilment and consumer notices.

LeadGenLab’s GEO and AEO services examine how product and brand information appears in AI search and shopping answers. Global teams can define a Korea-first or wider APAC audit through the project contact.

Official sources

• https://help.openai.com/en/articles/11128490-shopping-with-chatgpt-search
• https://help.openai.com/en/articles/6825453-chatgpt-release-notes
• https://help.openai.com/en/articles/20001052-using-library-to-manage-files-in-chatgpt

Checked October 10, 2026 against OpenAI Help Center and release notes. Availability and product presentation may vary by account, region, device and merchant.

FROM NEWS TO EXECUTION · LEADGENLAB

분석을 실제 실행으로 연결하려면

AI Search Lab은 공식 출처와 편집 기준에 따라 변화와 실행 기준을 설명합니다. 진단, 기술·콘텐츠 개선, 산업별 GEO·AEO 지원이 필요하다면 운영사 LeadGenLab의 관련 서비스와 상담 페이지에서 다음 단계를 확인하세요.

맞춤 솔루션GEO·AEO 대행 솔루션 →실무 인사이트Research & Insights →회사 소개Mission & Vision →Contact무료 진단·상담 신청 →