Should you fix missing AI mentions with a blog post or better product copy?
A missing AI mention can make a blog post seem like the obvious fix.
Someone asks ChatGPT or Perplexity for products like yours. Your store is not named. So you plan an article.
Sometimes that is the right move. But if the question is about a product you sell, start with the product page.
Your product page has one job: make the item clear enough to buy without extra research. A blog post has a different job. It helps someone understand a wider choice before they pick an item.
That difference matters more now that Shopify Catalog passes product information to AI sales channels. Shopify says product titles, descriptions, options, images, prices, availability, and other attributes help those channels understand products. Shopify’s product guidance makes the same practical point: accurate, detailed product information improves a product’s chance of appearing in AI search.
Start with the page that sells the item
Say your store sells waterproof dog coats.
A shopper might ask:
- What waterproof dog coat works for a small dog?
- Does this dog coat have a harness opening?
- Which dog coats work for heavy rain?
- What dog coats have reflective details?
Now look at a product page that says only: “A stylish coat for everyday walks.”
That copy does not answer much. The coat may be waterproof. It may have fleece lining, adjustable straps, reflective trim, and a harness opening. But perhaps those facts only appear in photos, a size chart image, or in somebody’s head.
The product page should state the useful facts in words:
- Outer fabric and lining
- Whether it is waterproof or water-resistant
- Harness opening
- Adjustable neck and chest straps
- Reflective details
- Available sizes and measurements
- Care instructions
This is not filler added for a machine. It is basic buying information.
It also gives AI systems a clearer record of what the item is. Shopify Catalog can carry product data to AI channels, and OpenAI’s commerce feed guidance includes product-level facts such as title, factual description, product URL, brand, image, price, and availability. OpenAI’s product feed specification is a useful reminder: a product needs facts, not just a nice phrase.
A blog post has a different job
A blog post works best when the question is broader than one product.
For the dog coat store, useful posts might cover:
- How to measure a dog for a raincoat
- Does a dog need a coat in wet weather?
- What makes a dog coat waterproof?
- How should a dog coat fit around a harness?
Those questions help a shopper work out what they need. The post can link to relevant products once they are ready.
But the post should not be the only place where a product fact lives.
If the harness opening is mentioned only in an article called “How to walk your dog in the rain,” a shopper still has to hunt for it on the product page. An AI assistant comparing items has the same problem.
Put core product facts on the product page. Use articles to explain decisions, trade-offs, and care.
ChatGPT’s shopping research can combine merchant product data with public product information and retail sources. It may surface product features, sizes, prices, trade-offs, and links to merchants. Those details can change, so the retailer’s own page still needs to be the clear source of truth. OpenAI explains the limits of shopping research here.
Do a one-minute product page check
Open one of your best-selling product pages.
Cover the photos if you can. Then answer these questions from the written page copy alone:
- What is this product?
- Who is it for?
- What is it made from?
- What problem does it solve?
- Which size, fit, dimension, or compatibility details matter?
- What makes it different from similar products?
- What might stop someone from buying it?
If you cannot answer those questions, the page needs more useful copy.
This often happens when key information lives only in an image:
- Furniture measurements in a diagram
- Ingredients on a label photo
- Device compatibility in a chart
- Fabric details in a close-up image
- Size guidance inside a graphic
Keep the image. Then write the important detail on the page too.
Alt text can describe what an image shows, especially a diagram or feature photo. But it is not a replacement for a product description. The shopper should not need an image description to learn the basic facts.
Use FAQs for the questions after the basics
FAQs are useful for practical objections and edge cases.
For a dog coat, that could include:
- Will this fit a standard harness?
- Can it be machine washed?
- Is the material suitable for sensitive skin?
- Does it fit a narrow-chested dog?
These answers belong well on the product page. They can also support FAQ structured data when the questions and answers are genuine.
Do not put every important fact into FAQs. A shopper should understand the item before reaching that section.
A simple page structure is enough:
1. A short description of what the product is. 2. Clear specifications. 3. Size, material, care, or compatibility details. 4. FAQs for questions that need more explanation.
Google recommends `Product` structured data for pages focused on a single product or its variants. It can help Google understand details such as product descriptions, price, and availability. But markup supports a clear page; it does not turn vague copy into useful product information. Google’s Product structured data guidance is clear that the page itself remains important.
Check that the product is available to be read
Better copy will not solve a product that AI shopping channels cannot access.
Descriva’s Agent view shows what AI shopping agents get from Shopify Catalog for each product. It flags products that are:
- Draft or unlisted
- Missing an image
- Priced at 0
- Not published to the Online Store
- Hidden from search engines
- On a store behind its password page
You can check several of these without any app.
In Shopify, open a product. Check its status, price, images, and sales channels. Then check its search engine listing settings.
Be careful with Unlisted. Shopify says unlisted products are hidden from AI discovery, but also from sitemaps, search engines, and your online store search. They may still be found elsewhere through crawling, indexing, or feeds depending on how they were shared before. Shopify explains product visibility for agentic storefronts here.
Use missing mentions to choose the right fix
Descriva’s AI Mentions asks the product questions shoppers ask and shows where your store is named, where it is absent, and which sources the answer cites.
On Starter, it checks ChatGPT and Perplexity. Pro adds Claude. Scale adds Gemini.
The missing question tells you what to fix.
If the question is “Which coats have reflective trim?” first check whether every relevant product page says “reflective trim.” If it does not, improve the product copy.
If the question is “How should a dog raincoat fit?” that may need a guide. Link that guide to the products it helps explain.
Descriva can draft product copy with specifications and FAQs, bulk alt text, and FAQ and schema packs through Fix autopilot. Every change waits for your approval. Fix buttons open Shopify’s own editors.
That review matters. Check every material, measurement, compatibility claim, and care instruction before publishing. A useful draft is not a substitute for knowing your own catalogue.
Build from product facts outward
Use this order:
1. Fix the product page. 2. Add specifications and practical FAQs. 3. Add accurate alt text where an image carries useful information. 4. Confirm the product is published and visible. 5. Write a blog post if the question needs a wider explanation.
We learned a version of this while building our own catalogue reader. Shopify’s `/products.json` returns no more than 250 products per page. A catalogue can look complete while quietly missing later pages. We had to handle pagination in our own `llms.txt` builder.
The same habit applies here: do not assume the information is present because you know it is somewhere. Open the live product page and read what it actually says.
Pick one product this week. Make it understandable without its photos. Then decide whether shoppers still need a separate guide.