Why product descriptions decide if you get found
Your product description is the main block of text on a product page. Google reads it to decide which searches the page can rank for. ChatGPT, Perplexity and Google's AI answers read it to work out what the product is and whether to mention it.
Most stores have the same gap: one-line descriptions, text copied from the manufacturer, or nothing at all. Writing 300 descriptions by hand rarely gets finished. AI can do that work, as long as it writes from real product data and you check it before it goes live.
What a good product description has in 2026
A good description gives shoppers and search engines the same thing: the answer first, then the facts. This is the shape we use on every product. The numbers match the notes on the right.
1A lightweight everyday runner that stays comfortable from the morning commute to the evening walk.
- 2Upper: breathable knit
- Sole: cushioned foam with rubber grip
- Weight: 260 g (US men's 9)
- Sizes: US 6 to 13
3Best for: daily walking, gym sessions and runs up to 10 km.
4Does it run true to size? Yes. If you are between sizes, go half a size up.
- One benefit sentence. What the product does for the buyer, in plain words. It is the line a search snippet or an AI answer can quote on its own.
- A spec list. Materials, sizes, weight, what is in the box. Numbers beat adjectives: "260 g" says more than "lightweight".
- Who it is for. One line on the use, so the page matches searches like "shoes for walking all day".
- Two to four buyer questions. Sizing, care, returns, compatibility. Take them from your support inbox and your reviews.
- SEO title and meta description. Keep the title to about 60 characters and the meta description to about 155, so Google is less likely to cut them off.
- Image alt text. A short, specific line per image: product name, colour, material. Google uses it to understand the image.
Write the facts once, reuse them everywhere. The spec list feeds the description, the SEO title, the alt text and your Product schema. Get the facts right and every other field gets easier.
Will Google penalize AI-written product descriptions?
No, not for being written by AI. Google's February 2023 guidance says appropriate use of AI or automation is not against its guidelines (Google Search Central). What Google acts on is scaled content abuse: many pages generated mainly to manipulate rankings, with little or no value to users, no matter how they were made (Google spam policies).
For a store the line is simple. A description written from that product's real details and checked by a person is useful content. Hundreds of near-identical paragraphs with only the product name swapped is the pattern the spam policy describes.
Google spots AI text and ranks it lower.
Google's guidance is about quality and intent, not about which tool wrote the words.
Specific facts help a page more than length.
150 words with real specs answers more than 600 words of filler.
One template for every product is fine.
Keep the structure the same, but every sentence has to come from that product's own details.
How wptaskify connects your AI to your store
wptaskify is an MCP server. MCP (Model Context Protocol) is an open standard that lets an AI assistant use tools. You connect your own Claude, ChatGPT or other MCP client once, and it can read and update your products through your store's official API.
wptaskifyMCP server- Product descriptions
- SEO title and meta
- Image alt text
- Preview, then batches of 25
Your AI writes the copy. wptaskify gives it the tools, and every tool below works the same way on both platforms:
| Task | WooCommerce tool | Shopify tool |
|---|---|---|
| Write one description | generate_product_description | shopify_generate_product_description |
| Fill the whole catalog | bulk_generate_product_descriptions | shopify_bulk_generate_descriptions |
| SEO title and meta | bulk_generate_product_seo | shopify_bulk_product_seo |
| Image alt text | bulk_product_alt_text | shopify_bulk_product_alt_text |
Three rules are built in. Every bulk tool shows a preview before it writes anything. Writes run in batches of at most 25 products, so you read one batch before the next. And by default the bulk tools only touch products with an empty description or missing alt text, so finished work is left alone. For alt text, your AI looks at each product image itself and writes a line for what it sees. See all 42 WooCommerce tools.
Step by step: descriptions for your whole catalog
- Connect your store
Add wptaskify to Claude or ChatGPT and connect WooCommerce or Shopify. It takes about two minutes and needs no code.
- Find the gaps
Ask: "List my products with no description." The bulk tool previews which products are empty before anything changes.
- Do one product first
Preview a single product. Check every fact against your own data and adjust the tone until it sounds like your brand.
- Give it your words
Name the tone (plain, friendly or technical) and two or three search terms per category, so descriptions match how people search.
- Run a batch of 25 and read it
Apply the first batch, spot-check the specs, then continue with the next batch. If the same problem keeps showing up, fix the instructions, not each product.
- Add SEO fields and alt text
Run the SEO title and meta tool, then the alt text tool, so every product page is complete.
- Check the results after four weeks
Use the checks in how to tell if it worked below.
A prompt you can copy
If you would rather write the instructions yourself, this prompt gives consistent results. Replace the parts in brackets.
Write a product description for [product name] using only these facts: [materials, sizes, weight, what is in the box, who it is for] 1. Start with one sentence that says what the product does for the buyer. 2. Add a bullet list of specifications. Use numbers where we have them. 3. Add one line that starts with "Best for:". 4. Add 3 short questions buyers ask about this kind of product, with answers based only on the facts above. 5. Write an SEO title under 60 characters and a meta description under 155 characters. 6. Tone: [plain / friendly / technical]. Do not invent specs. If a fact is missing, say so instead of guessing.
The last line matters most. It stops the AI from filling gaps with guesses, which is where wrong sizes, wrong materials and avoidable returns come from.
Before and after: one product
Here is the same product with a typical supplier description and with the structure above.
Classic Runner. Mesh sneaker. High quality and comfortable. Available in many sizes. Perfect for everyone.
No facts a shopper can check. No questions answered. Same words as other sellers.A lightweight everyday runner that stays comfortable from the morning commute to the evening walk.
- Upper: breathable knit
- Sole: cushioned foam with rubber grip
- Weight: 260 g (US men's 9)
- Sizes: US 6 to 13
Best for: daily walking, gym sessions and runs up to 10 km.
Does it run true to size? Yes. Between sizes, go half a size up.
And the SEO title and meta description written in the same pass, as they could look in Google:
A lightweight knit running shoe with a cushioned sole. 260 g, sizes US 6 to 13, free 30-day returns. Shop the Classic Runner in 6 colors.
Title: 54 characters · Meta description: 137 charactersThe new version is not much longer. It is more useful, because every line is either a fact a shopper can check or a question they would ask.
Where each field lives in Shopify and WooCommerce
Both platforms store the same kinds of text, in different places. Knowing where each one lives helps you check the result.
| Field | Shopify | WooCommerce |
|---|---|---|
| Main description | Description (the product body) | Product long description |
| Short summary | No built-in field | Product short description, shown near the price |
| SEO title | Search engine listing: page title | Your SEO plugin, such as Yoast SEO or Rank Math |
| Meta description | Search engine listing: meta description | Your SEO plugin |
| Image alt text | Alt text on each product image | Alt text in the media library |
| Product schema | Added by your theme | Added by WooCommerce, extended by SEO plugins |
On WooCommerce a product is a WordPress post, so the same SEO plugin that handles your blog handles your products. For a WooCommerce-only walkthrough, read how to write WooCommerce product descriptions with AI, and for the rest of store SEO, WooCommerce SEO beyond Yoast.
Product schema: what Google and AI read behind the page
Structured data is code on the page that states facts in a format machines read directly: name, price, currency, stock, brand and reviews. Google uses Product structured data for product rich results (Google Search Central). Check these on a few products after the rewrite:
- The name, image, description and brand in the schema match what the page shows.
- An Offer with
price,priceCurrencyandavailability, kept in sync with your stock. - A GTIN or MPN where the product has one.
aggregateRatingandreviewonly when the reviews are real and shown on the page.- Nothing marked up that a visitor cannot see. Google's structured data guidelines require this.
About FAQs: since August 2023, Google shows FAQ rich results only for well-known, authoritative government and health sites (Google Search Central). Keep the questions on your product pages anyway. Shoppers read them, and a clear question with a short answer is easy for an AI assistant to quote.
One more check for AI search: make sure your robots.txt does not block OAI-SearchBot, the crawler OpenAI uses to show sites in ChatGPT search results (OpenAI).
How to tell if it worked
- Nothing left empty. Run an audit and confirm no product still has an empty description, SEO title or alt text. The free AI SEO score checks a page for you.
- Search Console. In Performance, filter pages that contain
/product/(WooCommerce) or/products/(Shopify). Compare the 28 days after the change with the 28 days before it: impressions, clicks and the number of queries. - Rich Results Test. Test a few product URLs in Google's Rich Results Test and fix any Product schema errors.
- AI answers. Ask ChatGPT or Perplexity for the best product in your category and see whether your store comes up. The free AI visibility checker runs these questions for you.
- Store numbers. Watch add-to-cart rate and product questions to support on the pages you rewrote.
Give it about four weeks. Google has to recrawl the pages before anything changes, and that can take days or weeks. Judging the result the next morning tells you nothing.
Mistakes that make AI descriptions hurt
- Letting the AI guess. A wrong size or material leads to returns and bad reviews. Tell it to use only the facts you give it.
- The same paragraph on every product. Each product needs its own facts. Swapping in a new product name is not enough.
- Filler words. "High quality", "perfect for everyone" and "must-have" tell a shopper nothing.
- Blank SEO fields. Without a meta description, Google writes its own snippet from whatever text it finds.
- Publishing 500 products unread. Read a few from every batch before you move on.
The short version
Give the AI real facts. Ask for a benefit sentence, a spec list and buyer questions. Set the SEO title, meta description and alt text in the same pass, and publish in batches you have read. That is how a catalog of thin pages becomes one that Google can rank and AI tools can quote.
Sources
- Google Search Central: guidance about AI-generated content (February 2023)
- Google: spam policies, scaled content abuse
- Google: Product structured data
- Google: general structured data guidelines
- Google: changes to HowTo and FAQ rich results (August 2023)
- Google: title links and meta descriptions
- Google: image SEO and alt text
- OpenAI: crawlers and user agents
