{"id":29539,"date":"2025-06-11T11:31:11","date_gmt":"2025-06-11T09:31:11","guid":{"rendered":"https:\/\/wordlift.io\/blog\/en\/?p=29539"},"modified":"2025-10-02T11:06:25","modified_gmt":"2025-10-02T09:06:25","slug":"googles-ai-mode-product-pages","status":"publish","type":"post","link":"https:\/\/wordlift.io\/blog\/en\/googles-ai-mode-product-pages\/","title":{"rendered":"What does Google\u2019s AI Mode really want from your product page \u2014 and what exactly is Chunk Optimization?"},"content":{"rendered":"\n

To help brands prepare for AI-driven search, I built a simulator that models how AI systems like Google AI Mode<\/a>, AI Overviews, Perplexity, and ChatGPT retrieve and rank content.<\/p>\n\n\n\n

<\/p>\n\n\n\n

\"\"<\/figure>
\n

AI Mode<\/a> is Google\u2019s most significant shift in the commercial search experience to date.<\/strong> What is Google’s AI Mode? Currently available in the U.S., it transforms Google Search into a chat-like experience<\/a> powered by a generative AI agent. At its core, it combines Google\u2019s Gemini model with real-time search capabilities to produce conversational answers\u2014similar to what users see in Perplexity or ChatGPT Search. It\u2019s a multi-turn, interactive experience<\/strong> where users can refine their queries and receive synthesized, context-aware responses directly within Search.<\/p>\n\n\n\n

To demonstrate this, I recently ran an experiment on the product description page for the iconic Tiffany\u2019s Knot Wire Bangle<\/a><\/strong>.<\/p>\n<\/div><\/div>\n\n\n\n

<\/p>\n\n\n\n

Using a meta-prompt<\/strong> originally developed by Jean-Christophe Chouinard<\/a>, I asked to Google’s AI Mode:<\/p>\n\n\n\n

\n

\u201cShow me the raw tools requests for the following query [product-name]\u201d<\/p>\n<\/blockquote>\n\n\n\n

\"\"<\/figure>\n\n\n\n

The response included 5 key areas that supposedly<\/em><\/strong> are used to ground the response when a user asks information about the product:<\/p>\n\n\n\n

    \n
  1. What is the price?<\/li>\n\n\n\n
  2. What materials is it made from?<\/li>\n\n\n\n
  3. Are there customer reviews?<\/li>\n\n\n\n
  4. What sizes are available?<\/li>\n\n\n\n
  5. What is the history or inspiration behind the knot design (Collection history)?<\/li>\n<\/ol>\n\n\n\n

    Our simulator \u2014 which casts a broader net with 20 synthetic questions<\/strong> \u2014 accurately predicted all 5 key areas surfaced during the audit<\/strong>. We intentionally look at a broader set of sub-queries because these are inherently stochastic and influenced by personal user context<\/strong> (as highlighted by Elizabeth Reid, VP and Head of Search, in Google\u2019s blog<\/a>).<\/p>\n\n\n\n

    AI Mode is evolving to incorporate even more contextual signals. Soon, it will offer suggestions informed by your search history, and users will be able to opt in to connect other Google services\u2014starting with Gmail\u2014to further personalize the experience.<\/p>\n\n\n\n

    \"\"
    The tool is still under development, but it already reveals the potential sub-questions GoogleAI Mode (or similar system) might generate\u2014and evaluates their coverage based on the analysis of the text elements found on the page.<\/figcaption><\/figure>\n\n\n\n

    We then compared the page content against those 5 queries using a high semantic similarity threshold of 0.80<\/strong> to simulate a competitive AI retrieval environment.<\/p>\n\n\n\n

    The outcome:<\/p>\n\n\n\n