Descriptions and categories created by AI
Describing hundreds or thousands of products by hand is days of work and a real cost โ and in dropshipping there is the added problem of copied supplier descriptions that hurt your rankings. We deployed a solution in which AI writes unique, sales-oriented descriptions and automatically assigns products to the right categories and tags. The catalog fills itself โ consistently, for SEO and without duplicates.
How it works โ step by step
Product without a description
A new or imported product goes into the queue to be completed.
Product analysis
AI analyzes the name, parameters and store context to understand what it is describing.
A unique description
It writes an original, sales-oriented description for SEO โ not a copy from the supplier.
Categories and tags
It automatically assigns the product to the right categories and tags in the store.
Consistent language and tone
It keeps the language correct (PL) and the brand style uniform across the whole store.
Ready to sell
The product is complete and ready to publish without edits.
What you get
Hundreds of descriptions with no hire
What used to take days of copywriting now happens automatically, in the background.
Perfect for dropshipping
Unique content instead of copies from the supplier โ the end of the duplicate-content problem.
Accurate categories
Products land where customers look for them โ easier navigation and more sales.
Consistent brand style
One tone and quality of descriptions across the whole catalog, whatever the number of products.
Better visibility
Original, valuable content increases products' chances in Google.
Faster sales start
New stock is ready to sell right away, not after weeks of writing descriptions.
Details
Why it matters in large stores and dropshipping
With hundreds and thousands of products, describing by hand is unrealistic on cost. In dropshipping there is an added trap: every seller uses the same supplier descriptions, so Google treats them as duplicates and does not promote them. Automatically generated unique descriptions solve both problems at once โ scale and SEO.
How a description is created
AI gets the product context (name, features) and the store guidelines (tone, language, what to avoid) and creates an original text. The solution keeps the language and quality correct โ if something is off, it retries. As a result the descriptions are consistent and ready to publish without manual edits.
Categories and tags
Beyond the description, AI matches the product to the existing category tree and assigns accurate tags. The catalog organizes itself, and customers find what they are looking for faster โ which translates directly into sales.
An engine based on the Claude API, integrated with WooCommerce. It works product by product, with a "to-do" queue, retries and language validation, and reports the results by email.
Content generation (woo-ai-generator)
The engine fetches the list of "todo" products (get_products) and processes them one by one (generate_product) โ with a retry mechanism and language validation. It stores the status in the _wag_status meta, and sends a summary via its own HTML mailer (wp_mail) to the configured recipients.
Quality gating
A product without a photo is skipped (it comes back once a photo appears). It works with a passive visibility gate that reads _wag_status and the description length โ incomplete products do not reach the front end, so the customer never sees an "empty" product.