{
  "id": "E01",
  "url": "https://cartforagents.com/guides/spec-sold-product-guides",
  "title": "Spec-sold products: turn shopper requirements into category facts and rule-out conditions",
  "description": "Translate buying requirements into category fields, units, conditions, sources and rule-out conditions, with a worked humidifier worksheet and short contrasts for winter clothing, travel bags and cosmetics.",
  "answer": "Start from the requirements shoppers actually state in your category, not from keywords or a universal readiness score. For each requirement, decide which fact would prove it, its unit, the condition it applies under, where it comes from, and the result that rules a product out. A humidifier's noise level, a snowsuit's temperature rating and a backpack's carry-on size need different facts and conditions, so build the model one category at a time.",
  "take": "This is the core of how we think about AI shopping: products are chosen on requirements, so a brand's job is to publish the facts that prove or rule out each one. Build it one category at a time, starting with the questions customers already ask you.",
  "summary": "Start with what a shopper must know to rule a product in, or out.",
  "claims": [
    {
      "text": "For the search \"humidifier for dry winter air large room quiet top fill\", the 10 product descriptions returned stated: a decibel figure in 4 (none with setting, distance or method); a room area in 4 (none with test conditions); a runtime in 8 (2 naming the setting); top fill in 7. One more listing says \"Top Fill\" in its title only, so title-or-description is 8.",
      "source_title": "Cart for Agents: Receipts cited by our guides (Shopify agent-catalog snapshot, Oct 26, 2026)",
      "source_url": "https://cartforagents.com/research/agent-shopping-research#receipts-cited-by-our-guides",
      "source_date": "2026-10-26",
      "method": "observed",
      "limitation": "One search, one unpersonalized snapshot, description field only except where a title is named. Counting Shopify-generated metadata too, 6 of 10 records carried a decibel figure. Counts what was written, not how the products perform, and is not a category-wide rate."
    },
    {
      "text": "The listing whose description said it runs quietly enough to leave on overnight gave no decibel figure, and was one of two whose descriptions stated the setting for a runtime figure (\"on low\").",
      "source_title": "Cart for Agents: Receipts cited by our guides (Shopify agent-catalog snapshot, Oct 26, 2026)",
      "source_url": "https://cartforagents.com/research/agent-shopping-research#receipts-cited-by-our-guides",
      "source_date": "2026-10-26",
      "method": "observed",
      "limitation": "Single listing; shows that completeness varies by fact within one record, not a pattern."
    },
    {
      "text": "For \"toddler snowsuit rated for minus 30\", 2 of 10 listings stated a -30 °C figure. One listing's title says it \"protects to -30 °C\", and its description adds temperature ranges with the seller's activity guidance. The other listing's description rates only the hood's insulation to -30 °C.",
      "source_title": "Cart for Agents: Receipts cited by our guides (Shopify agent-catalog snapshot, Oct 26, 2026)",
      "source_url": "https://cartforagents.com/research/agent-shopping-research#receipts-cited-by-our-guides",
      "source_date": "2026-10-26",
      "method": "observed",
      "limitation": "One snapshot. Our first matcher missed both because it did not equate \"minus 30\" with \"-30 °C\"; this count is from a manual reread of the raw text. A stated rating is not a test of warmth or safety."
    },
    {
      "text": "For \"anti theft backpack for city travel\", one product description called the bag \"carry-on compliant\" without naming an airline or stating dimensions in that sentence.",
      "source_title": "Cart for Agents: Receipts cited by our guides (Shopify agent-catalog snapshot, Oct 26, 2026)",
      "source_url": "https://cartforagents.com/research/agent-shopping-research#receipts-cited-by-our-guides",
      "source_date": "2026-10-26",
      "method": "observed",
      "limitation": "The catalog record also carried generated dimensions; we did not check them against any airline limit or the manufacturer."
    },
    {
      "text": "In the E-GEO benchmark, 11 of 15 hand-written rewriting heuristics did worse on average than a simple rewrite prompt. The authors describe the optimized prompts as turning marketing prose into attribute-labelled bullets while preserving the product facts.",
      "source_title": "E-GEO: A Testbed for Generative Engine Optimization in E-Commerce (v2)",
      "source_url": "https://arxiv.org/html/2511.20867v2",
      "source_date": "2026-07-14",
      "method": "controlled-simulation",
      "limitation": "Authors report a controlled benchmark that re-ranks 10 already-retrieved Amazon listings per query with language models. It is not a deployed assistant, does not test retrieval, and does not measure sales."
    },
    {
      "text": "OpenAI says the factors ChatGPT weighs shift with the shopper's request; for example, a stated $30 budget makes it focus more on price.",
      "source_title": "OpenAI Help Center: Shopping with ChatGPT Search",
      "source_url": "https://help.openai.com/en/articles/11128490-shopping-with-chatgpt-search",
      "source_date": "2026-10-27",
      "method": "documented",
      "limitation": "Living help article checked on this date. Describes inputs, not weights or how often a product is shown."
    },
    {
      "text": "Shopify says some Global Catalog fields, including technical specifications, are generated or enriched by Shopify, marked Inferred, and may vary in accuracy.",
      "source_title": "Shopify: Global Catalog MCP, Inferred response fields",
      "source_url": "https://shopify.dev/docs/agents/catalog/global-catalog",
      "source_date": "2026-10-27",
      "method": "documented",
      "limitation": "Living developer documentation checked on this date; does not say how assistants use these fields."
    },
    {
      "text": "In a supplied-snippet simulation, an invented product was placed among 7 real catalog listings and each model picked 2 of the 8, so chance is 25%. A lip gloss with vague copy was picked in 0 sessions on every model; with its facts in the description it was picked in 168 of 180 sessions across three Claude models, 60 of 60 on GPT-6.1-Sol, 49 of 60 on Gemini 3.8 Flash and 53 of 60 on Gemini 3.1 Pro. A humidifier went from 3 of 180, 0 of 60, 6 of 60 and 3 of 54 with vague copy to 134 of 180, 20 of 60, 19 of 59 and 15 of 55 with its facts as a spec line. On GPT and Gemini the humidifier was picked in every session with a detailed shopper brief (20 of 20, 19 of 19, 15 of 15) and in 0 of 120 sessions with a one-line ask or no shopper.",
      "source_title": "Cart for Agents: Vague vs specific product facts, simulated picks by Claude, GPT and Gemini models",
      "source_url": "https://cartforagents.com/research/vague-specific-product-facts-simulation",
      "source_date": "2026-10-27",
      "method": "controlled-simulation",
      "limitation": "Invented products, snippets supplied by us, no page opening or tools, a forced choice of two with no \"none fit\" option, a detailed shopper brief written to fit the test product, and API models rather than the ChatGPT or Gemini shopping products. Not live visibility, rank or sales."
    }
  ],
  "sources": [
    {
      "title": "Cart for Agents: Receipts cited by our guides",
      "url": "https://cartforagents.com/research/agent-shopping-research#receipts-cited-by-our-guides",
      "accessed": "2026-10-27"
    },
    {
      "title": "E-GEO: A Testbed for Generative Engine Optimization in E-Commerce (v2)",
      "url": "https://arxiv.org/html/2511.20867v2",
      "accessed": "2026-10-27"
    },
    {
      "title": "OpenAI Help Center: Shopping with ChatGPT Search",
      "url": "https://help.openai.com/en/articles/11128490-shopping-with-chatgpt-search",
      "accessed": "2026-10-27"
    },
    {
      "title": "Shopify: Global Catalog MCP",
      "url": "https://shopify.dev/docs/agents/catalog/global-catalog",
      "accessed": "2026-10-27"
    },
    {
      "title": "Shopify Help Center: Shopify Catalog and product discovery for agentic storefronts",
      "url": "https://help.shopify.com/en/manual/online-sales-channels/agentic-storefronts/products",
      "accessed": "2026-10-27"
    },
    {
      "title": "Cart for Agents: Vague vs specific product facts, simulated picks by Claude, GPT and Gemini models",
      "url": "https://cartforagents.com/research/vague-specific-product-facts-simulation",
      "accessed": "2026-10-27"
    }
  ],
  "limitations": [
    "The category examples come from three searches we ran in Shopify's agent catalog on October 26, 2026, without personalization. We did not measure how often categories lack facts, and missing facts are not a ranking of product quality or business opportunity.",
    "The worksheet is a proposed method. We have not tested whether using it changes what shoppers buy or what any assistant recommends.",
    "Category standards for test conditions (how noise, coverage or temperature ratings should be measured) are not settled here. Where we do not name a standard, treat it as an open research gap, not as an answer.",
    "A temperature rating, noise figure or \"compliant\" claim describes the conditions it was tested under. It does not certify safety or fit in every situation.",
    "Store names and private catalog receipts are withheld."
  ],
  "related": [
    "https://cartforagents.com/guides/agent-ready-product-data",
    "https://cartforagents.com/guides/humidifier-field-guide",
    "https://cartforagents.com/research/quiet-humidifiers",
    "https://cartforagents.com/research/vague-specific-product-facts-simulation",
    "https://cartforagents.com/research/agent-shopping-research"
  ],
  "published": "2026-10-27",
  "updated": "2026-10-27",
  "checked": "2026-10-27",
  "kind": "guide"
}