How to Write Product Descriptions and Category Copy

The content problem I see most in ecommerce audits is manufacturer copy left unchanged across thousands of products, plus keyword paragraphs stacked under category pages that nobody reads. Both tell shoppers and search engines the same thing: this page offers nothing you cannot find elsewhere.

This post covers the template I use for product descriptions, what to do with manufacturer text, and where and how long category copy should be. For category structure, filters and technical setup, see the ecommerce SEO guide.

Short answer: a good product description answers the questions shoppers ask before buying: who it suits, why this product, specs, use and care, shipping and returns. Manufacturer text is only a starting point; enrich it with your own knowledge. Category copy should be a short intro that does not push the product grid down, optionally followed by a buying guide below the grid.

Why is copying manufacturer descriptions a problem?

When dozens of stores selling the same product use the same manufacturer text, Google is left with dozens of near-identical pages. For duplicate content like this, Google usually surfaces one or a few versions; which one often depends on the overall strength of the domain, and smaller stores lose that contest.

It is not about penalties; it is about distinctiveness. If your page offers nothing the competing page does not, there is no reason for it to rank. Product pages piling up under "Crawled – currently not indexed" in Search Console is a common symptom; I cover the diagnosis in crawled but not indexed pages.

Product description template: what to include

Every product category has its own questions, but the skeleton below works for most products. Gather information from customer service questions, return reasons and reviews; that is usually where the most valuable content sits.

A step often skipped is using the words shoppers actually use. You may write "cotton duvet set" while shoppers search "double duvet cover set percale". Find these terms through keyword research and place them naturally in the copy.

How to scale original copy across thousands of products

Writing long copy by hand for every product is unrealistic for most stores, so I set priorities:

AI tools can speed up drafting, but always check specs and dimensions against real product data. Google's position focuses on whether content helps users rather than how it was produced; details are in AI content and Google's policy.

Where should category copy go, and how long?

A category page's first job is to show products. A long intro that pushes products below the fold tires users and hurts conversion. There is no ideal word count; I recommend a two-part structure.

Instead of filler sentences repeating the category name, category copy should resolve the points where shoppers get stuck, such as "What is drop in running shoes, and which suits you?". To decide which query the copy should serve, use the intent-matching approach in the content strategy guide.

Product page checks before publishing

Once the copy is ready, check the rest of the page too. However good the description, missing structured data or wrong image text lowers the page's value.

• The title includes the product name shoppers search for, the brand and a distinguishing feature.
• The first sentence says who the product is for and why.
• Size, material and compatibility are in a verified table.
• Image alt text describes the image.
• Product structured data (price, stock, rating) matches the page.
• The page links to its category and to a guide.

You can build and test product structured data with the schema generator. For platform fields see the Shopify SEO checklist, and for marketplaces see Trendyol product page optimisation. If you want product and category copy written to a scalable plan, see content strategy services and ecommerce SEO consulting.

8+ years of hands-on SEO and GEO experience across 150+ projects in e-commerce, SaaS, healthcare and local business sectors. Writes based on real client data, not theory.