The supplements market is one of the most scrutinized by AI agents. ChatGPT Shopping and Perplexity seek to verify composition, per-active-ingredient dosage, and contraindications before recommending a product. A store that doesn't expose its dosages in mg/µg in schema.org Product (via additionalProperty) is structurally disadvantaged against a competitor that does. This checklist guides you through your industry's critical zones.
01 · The checklist
The data engines look for on a supplement product page
Excerpts from the sector checklist, ranked by weight: what blocks a recommendation when missing, then what reinforces trust. The full audit monitors 100+ signals.
Required content
Trust signals
Expected UX patterns
Personalized dosage calculator based on profile (weight, age, goal)
02 · Industry benchmarks
Reference values for a supplement page
Internal reference valuesPer-sector reference targets defined by Verity Score, not measured on a sample of stores.
03 · Required schema.org
The markup engines can read
Copy the structure and adapt the values to your pages.
"additionalProperty": [{"@type": "PropertyValue", "name": "Dosage", "value": "500", "unitText": "mg"}]"additionalProperty": [{"@type": "PropertyValue", "name": "Form", "value": "Capsules"}]04 · What leaders do
What the best-structured supplement brands do
Detailed composition with dosage per active ingredient
Clear delivery time in the schema
Return policy extractable from the JSON-LD
Customer reviews with schema.org markup
Complete Product schema with additionalProperty (dosage)
05 · Frequent questions
Your questions about AI engines
How do I expose dosage in structured data?
Add an additionalProperty of type PropertyValue with name='Dosage', value='500' and unitText='mg' in your schema.org Product. AI agents extract this data to compare products.
Are contraindications verified by AI?
Yes. Cautious AI systems look for visible contraindications, especially for pregnant/breastfeeding women and drug interactions. Expose them in both content and schema.
What impact does a nutritionist recommendation have?
An endorsement from a nutritionist or dietitian is major proof of expertise. Marked up, it is expert validation AI systems can read and verify.
What industry benchmarks does Verity Score use?
The supplements industry averages 8 structured fields, 12 AI-extractable claims, 3 trust signals in schema, and 4 content gaps. Your audit compares your store against these benchmarks.
Should the galenic form (capsules, powder, liquid) be in the schema?
Yes. Add an additionalProperty with name='Form' and value='Capsules' (or 'Powder', 'Liquid', 'Gummies'). AI agents use this data to filter results based on user preferences.
Do health-advice chatbots create merchant liability in 2026?
Yes. The Moffatt v. Air Canada (2024) ruling established full merchant liability for chatbot-provided information. New York Senate Bill S7263 (February 2026) creates explicit civil liability for chatbot health advice. Since August 2, 2026, Article 50 of the EU AI Act (Regulation (EU) 2024/1689) additionally requires clearly disclosing to users that they are interacting with an AI. Never allow a chatbot to recommend dosages or diagnose conditions : restrict it to verified product information only.
Which ChatGPT product feed fields matter most for a supplement?
Three fields from the OpenAI spec (Agentic Commerce developer docs, consulted August 13, 2026) change everything in this category : unit_pricing_measure and base_measure, which let you show the unit price (per capsule, per gram) and be compared fairly against a competitor with different packaging ; expiration_date, so the agent drops an expired batch instead of recommending it ; and warning plus warning_url for cautions (pregnancy, drug interactions).
Get your Supplements store audited
The Shopify app is available. It monitors 100+ signals across 39+ zones, produces an initial diagnostic in about 60 seconds, and prioritizes fixes for approval.
Free trial