September 19, 2026 · The lunalink.ai team

Should product answers go in descriptions, or is an FAQ enough for AI shopping agents?

Wooden letters spelling the word "QUESTIONS" on a cardboard background, providing a neutral copyspace. — Photo by Ann H on Pexels

A shopper asks an AI agent, “Will this jacket fit over a hoodie?” or “Does this charger work with my phone?”

Where should the answer live?

It is tempting to put it in an FAQ. FAQs are quick to add, and you may already have a good help page. But a product question needs to live with the product first.

For Shopify stores, product descriptions and structured product fields are the main source for product-level answers. FAQs have a different job. They help agents answer questions about your store as a whole.

That split matters when an AI agent is choosing between two products, comparing variants, or deciding whether an item matches a shopper’s need.

Put product facts on the product page

Shopify Catalog sends product-level fields to AI channels. These include the product title, description, options, images, price, availability, and other key attributes.

Shopify also lists product title, description, images, product type, vendor, collections, tags, barcode or GTIN, and variants among the details AI platforms can consider.

That means a product description is not just copy beneath the buy button. It is part of the product record an agent can use to understand what you sell.

Say your store sells a stainless steel water bottle in two sizes.

A useful product description could answer:

What material is it made from?
How much does each size hold?
Does it fit a standard cup holder?
Is it insulated?
Can it go in a dishwasher?
What is included?
Which size suits a school bag versus a day hike?

Those facts belong on the product page because they change whether the product is right for the shopper.

Do not make an agent infer them from product photos or a short title. Do not leave them only in a blog post, a review, or a general FAQ either.

Shopify’s guidance is direct here: product detail pages should include comprehensive product information, including specifications, comparison information, relevant keywords, sizing, materials, and care instructions. If some extra detail is useful to AI but makes the main product page harder for people to scan, Shopify suggests placing it lower on the page.

That is a practical middle ground. Your main description can stay readable. A lower section can hold dimensions, compatibility lists, care details, or a plain comparison table.

Use FAQs for questions about the store

Shopify Knowledge Base is a free first-party app available on all Shopify plans. It lets you create FAQs that AI shopping agents can use for store-level questions.

That is useful. It is just not a substitute for product data.

Shopify’s own FAQ examples focus on matters like:

Where you ship
Whether local pickup is available
Your return rules
Store credit rules
Supported languages
Customer account options

These answers apply across many products. They belong in a central place.

For example, an FAQ might answer:

Do you ship to Canada?

Or:

How long do returns take to process?

Or:

Can I collect my order in person?

Those are good Knowledge Base questions. Shopify recommends brief answers of one or two sentences for custom FAQs.

A question like “Is the 1-litre bottle dishwasher-safe?” is different. It refers to a specific product and may differ between products. Put that answer in the bottle’s description and relevant product fields.

If you also add it to a FAQ, treat the FAQ as a helper, not the only source. You do not want a shopper asking about one product and receiving a general answer that quietly does not apply to another.

A simple rule for deciding where an answer goes

Ask one question:

Would this answer change from one product or variant to another?

If yes, add it to the product page.

This includes:

Fit and sizing
Materials
Dimensions and weight
Ingredients
Compatibility
Technical specifications
What is included
Care instructions
Use cases and limits
Variant differences
Certifications or identifiers that apply to that product

If the answer applies to your whole store, it is a better FAQ candidate.

This includes:

Shipping destinations
Delivery expectations
Returns and exchanges
Pickup rules
Payment or account policies
Brand-wide service policies

There will be overlap. Shipping is a good example.

A store-wide FAQ can explain where you ship and your general return policy. Product data and structured fields can carry details that differ by item, such as price, availability, shipping information, return information, and variants.

Google makes a similar distinction in its product documentation. It recommends combining on-page `Product` structured data with a Merchant Center feed. The product data can include price, availability, reviews, shipping, returns, variants, and product identifiers. A standalone FAQ page does not provide the same structured product or feed fields.

Check one product in under a minute

Open one of your best-selling product pages. Then ask yourself the questions a shopper might ask before buying.

Try this list:

What is it made of?
What size is it, in useful units?
Who is it for?
What does it work with?
What does it not work with?
What differs between variants?
How should the buyer use, wash, charge, install, or store it?
What would make someone choose this product over another one in your catalogue?

Now look only at the product title, description, options, and visible product details.

Can you answer each question without opening another page?

If not, add the missing answer to the product record. Keep it factual. A short sentence is often enough.

For example, instead of writing:

Built for everyday adventure.

Write:

The 750 ml size fits most car cup holders. Hand wash the lid.

The first line may suit your brand voice. The second line answers a buying question.

Do not assume a catalogue read is complete

We learned this while building our own `llms.txt` builder.

Shopify’s `/products.json` endpoint returns at most 250 products per page. A catalogue read can look complete while quietly missing the rest. We had to handle pagination rather than assume the first response represented the whole store.

That is one reason a product-by-product check is useful. Missing answers are often uneven. Your newer products may be detailed, while older ones have only a title and two lines of copy. A collection can look well maintained from the storefront while important product facts are absent from the actual product data.

Shopify’s agentic storefront documentation also makes the hierarchy clear. Stores automatically expose `/agents.md`, `/llms.txt`, and `/llms-full.txt`, with store context such as policies and discovery endpoints. Those files do not replace Shopify Catalog, which remains the authoritative product-data feed.

Your product information still needs to be in the product information.

Where Descriva fits

Descriva checks this question using the catalogue Shopify serves to AI shopping agents.

It asks real shopper questions about your products, then checks each answer against your product text. Each result is marked:

Met: the product text supports the answer.
Partly: some support is present, but an important detail is missing.
Contradicted: the answer conflicts with the product text.
Not stated: the product text does not support the answer.

That gives you a concrete editing list. You can see whether a question needs a line in a product description, a clearer variant option, or a store-level FAQ instead.

Descriva includes one free assessment per store, along with its SEO toolbox.

You do not need to rewrite every product description at once. Start with products that drive the most questions, have several variants, or need a buyer to understand fit, compatibility, or care before purchasing.

Sources

shopifyai searchproduct descriptionsfaqs