How Stanley 1913 Restructures Content for AI Search and LLM Recommendations

American brand Stanley 1913 is restructuring its content strategy for AI search: adding FAQ to product cards, care instructions, and detailed purchase occasion descriptions — everything large language models need to recommend products in ChatGPT or Perplexity. According to Pew Research Center, 42% of American adults already use AI chatbots for information searches, yet most brands' content is still built for human visual perception rather than LLM parsing.

Why visual content stops working in AI search

Stanley 1913 built its recognition on creator content: visual campaigns for Mother's Day, Teacher Appreciation Week, collaborations with influencers. Humans extract meaning from images, faces, and tone — language models can't do that. LLMs aggregate text data from open pages: product descriptions, articles, reviews, FAQ. Without detailed text, a brand simply won't appear in ChatGPT responses, even with millions of social media reach.

Stanley 1913's Chief Brand Officer Kate Ridley explains: "We created lots of visuals around holidays, but didn't explain why a thermal mug makes a good gift for a nurse or teacher. Now we understand: for LLMs, that's critical." The brand started embedding answers to questions people ask AI assistants directly into product cards: "what to gift for Mother's Day," "how to choose a fitness thermos," "stainless steel cookware care."

How Stanley 1913 adapts content for language models

The company launched a cross-functional task force: content, SEO, e-commerce, technology, PR, and marketing. The goal isn't just to add text to the website, but to build a unified product data system that LLMs can parse from any source. Ridley emphasizes: "We're building guides and metrics specifically for LLMs so all departments work in sync."

At the infrastructure level, Stanley 1913 is testing Shopify and Google's Universal Commerce Protocol — a protocol that pulls product catalogs directly into chatbot conversations. The brand uses Schema.org structured markup so data reads consistently for both traditional search engines and AI agents. Working with the Yotpo Discovery platform, the team tracks how often Stanley 1913 products appear in LLM responses and are cited as sources.

"Stanley 1913 brilliantly recaptured attention through real creators and influencers. Now brands need to view AI platforms as a new type of influencer — places where consumers discover brands, learn about them, and make purchase decisions," — Debra Aho Williamson, founder of Sonata Insights.

Working with external sources and earned media

LLMs rely not only on brand websites but also on third-party publications: media reviews, partner content, customer feedback. Stanley 1913 tracks which platforms AI search engines consider authoritative and directs PR efforts there. Cultural campaigns — for example, a recent partnership with singer Kacey Musgraves — generate press mentions and discussions beyond owned channels, increasing the chances of appearing in LLM results.

Ridley clarifies: "We're not abandoning social media for AI platforms. We need both to provide maximum product information. But for LLMs, information must be in explicit text form — otherwise the model simply can't work with it." Stanley 1913's website traffic grew 35.5% year-over-year, reaching 6.6 million visits in July 2026 according to Similarweb — and as audience grows, so does the number of touchpoints where consumers can discover the brand.

What this means for the Russian market and brands

In Russia, the share of search queries through AI assistants is still lower than in the US, but the trend is accelerating: Yandex integrated YandexGPT into search, users increasingly turn to Alice and third-party chatbots for recommendations. Russian brands face the same problem: visual content for Stories and Reels doesn't contain structured data that language models need.

Practical takeaway: if a product card on a marketplace or corporate website doesn't answer the questions "what for," "who is it for," "how to use," brands risk being filtered out of AI recommendations. This is especially critical for competitive categories — cosmetics, electronics, sports goods — where consumers choose between dozens of similar offerings.

Checklist: how to adapt content for AI search

• Product card audit: check if descriptions answer typical questions — "how to choose," "for what occasion," "what's different from competitors." If the description is under 300 characters and just lists specs, it needs expansion.
• Add product-level FAQ: place 3–5 questions with answers directly in the card. Phrase questions the way people ask them in search: "can it be washed in the dishwasher," "suitable for hot drinks," "what volume to choose for an office."
• Structured markup: implement Schema.org types: Product, FAQPage, HowTo. This helps search engines and LLMs correctly interpret page data.
• Occasion-driven content: create guides and collections for specific use scenarios — "gifts for colleagues," "thermoses for hiking," "dishes for picnics." Each guide should contain detailed descriptions, not just photos and product lists.
• Monitor AI mentions: regularly check how chatbots answer questions in your category. If your brand doesn't appear in responses, analyze which sources the model cites and work on your presence there.
• Department synchronization: content, SEO, e-commerce, and PR should work by unified guidelines. Create an internal document with product description requirements and update it with AI optimization tasks.

Metrics and measuring results

Stanley 1913 uses partner tools to track how often it's mentioned in LLM responses and how its products are cited. Russia doesn't yet have ready-made analytics platforms for GEO (Generative Engine Optimization), but you can measure indirect metrics: growth in branded searches on Yandex after publishing expanded descriptions, changes in website positions in "quick answer" blocks, increased traffic from chatbots in web analytics.

It's important to understand: proving commercial impact from AI optimization is harder than calculating ROI from traditional contextual advertising. Brands invest in tools to assess LLM result visibility, but establishing direct sales correlation is still difficult. Stanley 1913 is betting on a long-term strategy: building data infrastructure now to stay visible as AI search's market share grows.

Frequently asked questions

What is content optimization for AI search and LLMs

It's adapting texts on your website, in product cards, and in external publications to meet large language models' requirements: adding FAQ, detailed usage scenarios, structured markup, and answers to natural user questions. The goal is to appear in responses from ChatGPT, Perplexity, YandexGPT, and other AI assistants that aggregate information from open sources.

How to measure content effectiveness for language models

You can track how often your brand and products are mentioned in LLM responses using specialized platforms (e.g., Yotpo Discovery), monitor growth in branded searches on traditional search engines after publishing expanded descriptions, analyze chatbot traffic in web analytics, and track positions in "quick answer" blocks. Direct sales correlation is hard to establish, so brands focus on visibility metrics.

Do we need to abandon visual content for text

No, visual content remains important for social media and creator campaigns. The task is to supplement it with text data that language models can parse: purchase occasion descriptions, instructions, FAQ. Stanley 1913 continues working with influencers while simultaneously enriching product cards with information for AI search.