# Start with two AI shopping surfaces > A bounded case for measuring ChatGPT and Google's Gemini and AI Mode first, then deciding which other AI-shopping surfaces belong in the next test. - Canonical: https://research.caeliai.com/blog/two-ai-shopping-platforms-that-matter - Markdown: https://research.caeliai.com/blog/two-ai-shopping-platforms-that-matter.md - Publisher: Caeliai (https://caeliai.com), independent AI commerce research and advisory practice - Official LinkedIn: https://www.linkedin.com/company/caeliai/ - Contact: contact@caeliai.com --- For ecommerce, start with **ChatGPT** and **Google’s Gemini / AI Mode**. Then decide which other surfaces belong in the next test. **Figure:** Fig. 01 · Platform scale First-party figures · 2025-26 - **800M+** - Weekly ChatGPT users researching and buying · OpenAI, 2026 - **2.5B+** - ChatGPT messages per day · OpenAI, 2025 - **900M+** - Monthly Gemini app users · Alphabet, June 2026 - **5T+** - Google searches per year · Google, 2025 ## A useful surface connects a question to a product and then to a measurable store path. **01 People** ### Mass consumer usage. Not a niche research audience. **02 Products** ### Real product answers. Visual recommendations, comparisons, prices, and merchant links. **03 Catalog** ### A product-data path. Shopify Catalog, Google & YouTube, or direct merchant feeds. **04 Sales path** ### A measurable store path. A store visit or checkout that can be measured separately from a completed purchase. ## The two ecosystems. **Priority 01 · ChatGPT** ### The shopping conversation. - Visual product discovery and comparison. - Shopping research for complex buying decisions. - Shopify Catalog already integrated. - Merchant and product paths back to the store. **Priority 02 · Google: Gemini + AI Mode** ### The search-to-shopping ecosystem. - 900M+ Gemini app users. - 1B+ monthly AI Mode users. - More than 5T Google searches each year. - Shopify products supplied through Google & YouTube. *Caption:* The commerce rails under each ecosystem *Caption:* Both ecosystems document a path from assistant to product data to merchant destination. Other assistants may differ by market and feature set. This is a prioritization argument, not a claim that Claude, Perplexity, Copilot, or every other answer engine has zero value. For most ecommerce brands, those are second-pass tests after the two largest consumer shopping surfaces are measured. ## Shopify shows downstream outcomes. It does not show the full path. **Shopify can show** - AI-channel sales and orders - Sessions and conversion - Connected channels - Catalog sync status - A directional Catalog preview **The missing answers** - Which shopping prompts return the brand? - Which competitors appear instead? - Which products never surface? - Where is the returned destination? - Which content and catalog fixes matter first? *Caption:* Shopify’s agentic dashboard · public merchant example · July 2026 [Image: Shopify agentic storefronts dashboard showing AI-channel sales, orders, and sessions] *Caption:* The scoreboard: sales, orders, sessions by AI channel. It reports outcomes. It cannot explain them. *Caption:* Shopify’s own caveat on its ranking preview · July 2026 [Image: Shopify documentation noting that AI platforms re-rank catalog preview results with their own logic] *Caption:* Shopify warns that AI platforms re-rank its Catalog preview with their own logic. The preview is directional, not a record of what customers saw. > Shopify reports downstream outcomes. A diagnostic records the path that precedes them. ## Measure the same five questions across both platforms. **01** ### Does the brand appear? **02** ### Which products appear? **03** ### Who appears instead? **04** ### Where does the returned link lead? **05** ### Can downstream traffic be measured? *Caption:* The path that questions 03 and 04 expose ## Measure the AI shopping path. In five business days, document where the path appears, where destination quality changes, and the first three changes worth making. Includes 100 brand- and product-focused checks across ChatGPT and Gemini, competitor and catalog analysis, Shopify/GA4 metrics, and a concise action plan. One fixed fee, fully credited toward implementation within 14 days. **Primary sources** [OpenAI: retail and ChatGPT usage](https://openai.com/solutions/industries/retail/) · [OpenAI: Powering Product Discovery in ChatGPT](https://openai.com/index/powering-product-discovery-in-chatgpt/) · [OpenAI: 2.5B+ messages per day](https://cdn.openai.com/global-affairs/be0fe9e0-eb97-43d1-9614-99f2bd948bcc/OpenAI_Productivity-Note_Jul-2025.pdf) · [Alphabet: June 2026 investor presentation](https://blog.google/alphabet/investor-presentation-june-2026/) · [Google: 5T+ searches annually](https://blog.google/products/ads-commerce/google-search-ai-brand-discovery/) · [Shopify: agentic storefronts](https://help.shopify.com/en/manual/online-sales-channels/agentic-storefronts) · [Shopify: agentic reporting and ranking preview](https://help.shopify.com/en/manual/online-sales-channels/agentic-storefronts/agentic-home) Audience and message/search figures use different reporting periods and are not directly additive. Neither OpenAI nor Google publishes a complete shopping-prompt total. The figures above establish reach; the commerce integrations establish relevance.