# Win the eight questions that close the sale > Canonical HTML: https://verityscore.io/en/product/product-answers/ > Markdown alternate: https://verityscore.io/en/product/product-answers.md > Last updated: 2026-08-19 > Content synchronized with canonical HTML. On every product, Verity finds the buyer questions your pages already win, and hands you the exact facts that would win the rest. ## AI Summary Verity Score checks 8 AI Buyer Score dimensions on every Shopify product: the questions a buyer puts to an assistant before purchasing. For each product it separates questions the merchant own published evidence can answer today from questions waiting on merchant data, names the missing facts in plain language rather than error codes, and publishes approved answers back as public product facts an assistant can read. ## Key points - **What questions do AI assistants ask about a product**: Verity Score checks 8 buyer dimensions on every Shopify product and separates the questions the merchant published evidence can answer today from those waiting on merchant data. Source: https://verityscore.io/en/product/product-answers/ - **Product-level AI answer coverage for Shopify**: Coverage is measured per product rather than per site, missing facts are named in plain language, and approved answers are published back as public product facts. Source: https://verityscore.io/en/product/product-answers/ ## An assistant will not guess, it will move on When a buyer asks whether something fits, ships in time, or suits a constraint, an assistant needs a fact it can point to. If your page does not carry it, the assistant does not infer it. It recommends a shop whose page does. That is why coverage is measured per product rather than per site. A store can score well overall and still lose every answer on its best sellers, because the facts that decide those specific purchases are the ones missing. Verity names those facts. Not a schema warning, the actual sentence a buyer needs. ## Key Proof Points - **8 buyer dimensions checked**: The 8 AI Buyer Score dimensions, run on every product, not sampled. - **2 states, deliberately**: Answerable now, or waiting on merchant data. No third bucket to hide behind. - **0 invented facts**: Where evidence is missing, Verity asks you for it instead of filling the gap. ## Per product, not per site A site-wide score hides the products that matter. Coverage is computed on each product, so your best sellers are judged on their own evidence. - Every product in the plan checked, not a sample. - Sortable by what a gap is likely to cost you. - Variants considered, because a buyer asks about the one they want. ## Two states, and no third A question is either answerable from what you have published, or it is waiting on a fact only you hold. There is no partial credit, because a buyer does not get a partial answer. - Answerable now: the evidence already exists on your pages. - Waiting on you: a real fact is missing and Verity says which. - No inference, no filler, no assumed values. ## Missing facts in plain language Not a schema error code. The sentence a buyer needs, phrased as a question you can answer in a few words. - Written for a merchant, not for a developer. - Grouped, so one answer can close the gap on many products. - Reusable across your catalog once supplied. ## Published where an assistant reads it Once approved, the fact does not just sit in a report. It is published as public product data on your storefront, which is what an assistant actually reads. - Published as public product facts, not stored in a dashboard. - Structured so an assistant can parse and attribute it. - Reversible for 30 days, like every Verity change. ## How coverage is closed - Verity checks the 8 buyer dimensions on each product against the evidence already published on your pages. - Each question lands in one of two states: answerable now, or waiting on a fact only you hold. - Missing facts are named in plain language and grouped, so one answer can close the gap on many products at once. - You supply or confirm the fact, review the draft, and approve it. - The fact is published as public product data, and the next audit re-checks the same eight questions. ## What this does not do - Verity will not invent a product fact. If the evidence does not exist, the question stays marked as waiting on you rather than being quietly filled. - Coverage measures what your pages can answer, not whether an assistant chose to use it. The shopper tests measure that. - The free plan covers up to 25 products, enough to see the pattern across a catalog but not to close it. ## FAQ ### What are the eight questions? They are the 8 AI Buyer Score dimensions: the recurring things a buyer needs settled before paying, such as fit and suitability, delivery, returns and the evidence that the claim holds. They are checked on every product in your plan. ### Why not just add schema markup? Markup declares a fact, it does not create one. If your page never states the material, the shipping window or who the product suits, no amount of structured data gives an assistant something to quote. Verity finds the missing fact first, then publishes it correctly. ### Can it fill the gaps automatically? Only where the answer already exists somewhere in your store, and then only after you approve it. Where the fact is genuinely absent, Verity asks you rather than inventing it. ### Do I have to answer product by product? No. Missing facts are grouped, so a single answer about your shipping policy or your sizing can close the same gap across a whole collection. ## Sources - [Google: Product structured data](https://developers.google.com/search/docs/appearance/structured-data/product) - [schema.org Product](https://schema.org/Product) ## Preferred Citation When citing this page, use the canonical HTML URL: https://verityscore.io/en/product/product-answers/