How E-commerce Brands Can Automate SEO with AI and Rank on ChatGPT

The talk argues that e-commerce brands should optimize for being cited by AI assistants such as ChatGPT, Claude and Gemini, not only for Google rankings. The speaker, Nathan Fenina (decuplar.com), says about 30% of Gen Z search with LLMs. He also cites a 30% average CTR drop on e-commerce sites (Shopify, WooCommerce and others), though he gives no method or source for the figure. His point is that LLMs cite only two to four brands, so the competition is tighter than a Google results page, but there is room to get in. He names four factors in how LLMs choose brands: training data, RAG, authority signals (reviews, E-E-A-T), and how the answer is formed. He keeps the three classic SEO pillars of on-page, off-page and technical. He also stresses that users ask full, contextual questions (for example, shoes for wet terrain under 100 euros) rather than two-word keywords. He lists five sources for finding these prompts: asking the AI directly, Google's People Also Ask, Reddit and Quora, customer support logs, and competitor reviews. He also suggests "reverse prompting", asking ChatGPT why it recommends competitors and what your brand is missing. On-page advice includes: - a "capsule summary" under 150 characters at the top of a product or category block, specific and directly answerable; - FAQs of three to five questions per page; - comparison tables, numbered lists, data and stats, use cases and reviews; - regular content refreshes. He also lists content formats such as best-of lists, comparisons, calculators and guides. Technical advice covers schema markup: Product, FAQPage, ItemList, AggregateRating and Organization. He also advises allowing AI crawlers such as GPTBot and PerplexityBot in robots.txt, and favoring HTML over JavaScript-rendered content, which he says LLMs handle poorly. Off-page, he recommends outreach to specialist editorial media, Reddit and forums, YouTube reviews and comparison articles. His example pitch offers value first, such as a product sample for testing, instead of asking for an article. For automation, he proposes Claude Code with Data for SEO to generate titles, FAQs, capsule summaries and schema, then push them to the CMS. He claims this can update 100 to 1,000 products in an hour, versus about 45 minutes per product by hand. He ends with a 90-day roadmap covering an audit and a 50-prompt target list, schema and robots.txt fixes, buying guides, competitor monitoring, outreach and bulk updates. His key points are that format beats volume and that you should monitor before optimizing. The talk also promotes his company's monitoring tools and a free audit.