How to Get Your WooCommerce Products Recommended by ChatGPT

When a shopper asks ChatGPT "what's a good waterproof dog coat for a small terrier?", it doesn't hand back ten blue links. It names a few products from a few shops. The question for your WooCommerce store is simple: does it have what ChatGPT needs to name yours?

This guide walks through what AI shopping assistants actually read, what we found when we checked 620 UK WooCommerce stores in October 2026, and which fixes you can make yourself, step by step. No developer required for most of it.

What we checked, and how

Between 9 and 11 October 2026 we ran a live audit on UK online shops. For each store we read the product list from WooCommerce's public Store API (up to 1,000 products), then opened up to 25 product pages spread across the catalogue and looked at what's actually in the HTML: product data, meta descriptions, descriptions and names.

This article uses the 620 WooCommerce stores where that audit was reliable (at least 20 product pages read, not blocked). Together that's 340,749 products and 15,385 product pages. The stores are independent UK shops from our own outreach list, mostly 10 to 2,500 products, so think "small and mid-sized UK WooCommerce", not "every shop on the internet". Our full write-up with Shopify numbers is in the UK data report.

Here's what stood out for WooCommerce:

• 22.6% of stores had no product data AI can read on any product page we opened, and 25.6% had it on fewer than half
• 44.2% had thin descriptions (under 150 characters) on at least 15% of their products, and 22.4% on at least half
• 47.7% of all products had no short description
• 50.1% of product pages had no meta description
• 40.5% of stores had at least one product named with an internal stock code

None of these are exotic problems. They're the normal state of a busy shop where products were added in a hurry. But each one makes it harder for an AI assistant to understand, trust and recommend a product.

How ChatGPT decides which products to show

ChatGPT, Google AI Mode and Perplexity all work roughly the same way when someone asks a shopping question. They look for products that clearly match the question, then pick the ones they can describe with confidence.

"Clearly match" comes from words: the product name, the description, tags and specs. If your listing for a dog coat says "DC-114 Navy M" and two lines of text, the AI has very little to match against "waterproof coat for a small terrier".

"With confidence" comes from facts it can check: a price, whether it's in stock, what it's made of. These come from your product pages (through product data AI can read, often called structured data or Schema.org markup) and from product feeds, like the one you might already send to Google Merchant Center.

So there are two layers to get right: the words, and the readable facts. If you want the bigger picture first, our guide on what AI commerce is explains how this shift works.

Step 1: Give every product a clear name

Start with names, because they're the first thing an AI matches on. Across our WooCommerce sample, 40.5% of stores had at least one product whose name started with an internal code, like "SKU-2231 Oak Shelf" or "BX40 Black".

Codes are useful for your warehouse, not for shoppers. A good AI-friendly name says what the thing is, plus the one or two details people search by:

• Instead of "DC-114 Navy M", try "Waterproof Dog Coat with Fleece Lining - Navy, Medium"
• Instead of "Candle 03", try "Fig and Cedar Soy Candle, 200g"

Keep the SKU in the SKU field, where it belongs. WooCommerce has one for exactly this reason.

You can do this yourself: Products > All Products, sort by name, and fix the ones that start with codes. Quick Edit lets you rename without opening each product.

Step 2: Write descriptions AI can quote

This is the biggest gap we found. In 44.2% of WooCommerce stores, at least 15% of products had descriptions under 150 characters. Across all 340,749 products we read, 28.3% were that thin.

150 characters is about one sentence. That's not enough for an AI to answer a follow-up question like "is it machine washable?" or "will it fit a Border Terrier?". When it can't answer, it tends to recommend something else.

A description that works for AI is simply a description that works for a careful shopper:

1. 1What it is and who it's for in the first sentence
2. 2What it's made of, size and fit, and anything people compare (weight, capacity, materials)
3. 3How it's used: the situations, the problems it solves
4. 4Care and practical details: washing, assembly, what's in the box

Aim for a few short paragraphs. Write the way you'd explain it to a customer in your shop, not a list of keywords.

You can do this yourself, and for your best sellers you probably should. The hard part is doing it for hundreds of products. This is where a tool earns its keep (more on that below).

Step 3: Fill in the short description

WooCommerce has two description fields. The short description usually shows next to the price and the "Add to basket" button, and many themes and plugins also use it as a summary.

In our sample, 47.7% of WooCommerce products had no short description at all, and in 42.6% of stores at least half the products were missing one. A one or two sentence summary ("A lightweight waterproof coat for small dogs, with a fleece lining and reflective trim") gives AI a neat, quotable summary of the product.

You can do this yourself: it's a quick win, one or two sentences per product.

Step 4: Check your product pages have product data

When an AI tool opens a product page, it looks for a block of machine-readable product data: the name, price, currency, stock status, brand and sometimes reviews. On WooCommerce this is usually Schema.org Product markup in a format called JSON-LD.

Here's the surprise: WooCommerce itself normally outputs basic product data on product pages. Yet 22.6% of the WooCommerce stores we checked had none on any product page we opened. When that happens, something is usually removing or replacing it: a theme, a page builder, or an SEO or speed plugin.

How to check your own store:

1. 1Open one of your product pages
2. 2Paste the URL into Google's Rich Results Test (search for it, it's free)
3. 3Look for a "Product" item with a price and availability

If there's nothing, check your theme and plugins first. SEO plugins such as Yoast (with its WooCommerce add-on) or Rank Math can add or extend product data. If you use a page builder for product pages, test a page built the default WooCommerce way to see if the builder is the cause.

This one is in your hands: it lives in your theme and plugins. Pickastor doesn't add markup to your pages, but the free AI Score checks your product pages and tells you whether the data is there.

Step 5: Add meta descriptions to product pages

Half of the WooCommerce product pages we opened (50.1%) had no meta description, and in 39.0% of stores none of the checked pages had one. Unlike Shopify, WooCommerce doesn't fill this in for you; it usually comes from an SEO plugin.

A meta description is the short summary search engines, and the AI tools built on top of them, often use when they describe a page. If you have Yoast or Rank Math installed, each product has a field for it. If you have many products, both plugins let you set a template that pulls in the product name and short description, which is a decent start.

You can do this yourself with your SEO plugin.

Step 6: Let AI crawlers in

None of the above matters if AI tools can't reach your store. Two settings to check:

• robots.txt (visit yourstore.com/robots.txt). Make sure it doesn't block AI crawlers such as OAI-SearchBot and GPTBot (OpenAI), PerplexityBot and Googlebot. Some security plugins and hosts block "bots" by default
• Cloudflare or your host's firewall. "Bot fight" modes can block AI crawlers even when robots.txt allows them

You may also have heard of llms.txt, a newer optional file that gives AI tools a short summary of your site. It's cheap to add and harmless, but it's not a replacement for the basics above, and not every AI platform uses it yet.

Step 7: Send a product feed

Product feeds are the most direct way to hand your whole catalogue to shopping platforms in one go. For WooCommerce, the free Google for WooCommerce extension connects your store to Google Merchant Center, which feeds Google Shopping and Google's AI features. OpenAI has also opened a way for merchants to share product feeds with ChatGPT; check chatgpt.com/merchants for the current status in your country.

A feed is only as good as the product data in it. A feed full of code names and one-line descriptions just delivers the same weak data faster.

What you can do yourself, and where a tool helps

Everything above is doable by hand. The honest split looks like this:

Quick to do yourself (an afternoon):

• Check product data with the Rich Results Test, and fix theme or plugin settings
• Check robots.txt and your firewall for AI crawlers
• Set up Google for WooCommerce and Merchant Center
• Add meta description templates in your SEO plugin

Slow to do by hand (it scales with your catalogue):

• Renaming products that use stock codes
• Rewriting thin descriptions
• Writing hundreds of short descriptions
• Adding tags and the phrases shoppers actually type

The second list is where most of the time goes. In our sample the median WooCommerce store had around 500 products. At 15 to 20 minutes per product, rewriting even a quarter of them is a few weeks of evenings.

How Pickastor fits in

Pickastor does the slow list. You connect your WooCommerce store with API keys, pick the products, and Pickastor rewrites each one: a clearer name, a detailed description, a short description, tags, product attributes (material, size, colour) and the natural search phrases shoppers use. You see every change side by side, and nothing changes in your store until you push it. On WooCommerce, the push updates the name, description, short description and tags.

Pickastor also publishes an AI product feed for your store (XML in Google Merchant Center format, plus JSON) with every product you optimize.

What it doesn't do: it doesn't add Schema.org markup to your pages, change your robots.txt or llms.txt, or edit image alt text, brands, reviews or barcodes. Those stay in your theme, plugins and settings, which is why the steps above are worth doing either way.

The first 3 products are free, and after that it's a one-off payment by product count, no subscription. See pricing, or start with the free AI Score to see where your store stands.

A simple order of work

If you only have an hour this week:

1. 1Run the free AI Score on your store
2. 2Check one product page in Google's Rich Results Test
3. 3Check yourstore.com/robots.txt for blocked AI crawlers
4. 4Rename your five best sellers so the name says what the product is
5. 5Rewrite the description of your single best seller, properly

Then work down the catalogue, best sellers first. AI recommendations behave a lot like SEO: changes take weeks to show up, not hours, and nobody can promise a specific result. But a store with clear names, real descriptions and readable product data gives ChatGPT every reason to pick it.