A store with eighty brands has hundreds of landing pages. Every brand page, every category and every combination of the two.
Category SEO is the work of giving those pages copy that matches what is actually in them today. Here is what that involves and what it delivers.
What it is
Each category page gets three things: an SEO title, a meta description and the copy on the page itself. Per language.
Not once, but again when the collection changes. The copy is an output of your catalog rather than a separate document sitting next to it and quietly going out of date.
Where the copy comes from
Three sources, and all three are needed.
The editorial brief per category: the angle, the brands that belong in it, the questions customers ask. Your own SEO agency can deliver it as a spreadsheet and it is imported as is.
The catalog facts: which products, brands and subcategories are actually in it right now.
Your store's format document: structure, length, linking rules and banned words, with its own version per language.
How you run it with AI in Automated Commerce
The agent does not write in isolation from your store. It receives the brief, the catalog facts and the format as input, and returns a proposal per category and per language.
- Import the briefs. Your SEO agency's sheet is turned into one brief per category inside the system.
- Set the format. Structure, length, linking rules and banned words, with its own version per language.
- Pick the categories and the languages and start the run. Up to five languages per run.
- The agent writes, and checks every text against the format's checklist before staging it.
- You approve on the board. Only then does the copy reach the collection and move on to Shopify.
What the AI does not do here matters just as much: it invents no brands, no categories and no product attributes. If a fact is not in your catalog or in the brief, it does not appear in the copy.
What comes out per category
Copy with a fixed structure, written per language rather than translated afterwards. That keeps one market's house terms out of another market's pages.
Every brand and every neighbouring category the copy names is linked to a collection that actually exists. A category you do not carry is never named.
At the bottom sit the questions from the brief, answered on the page itself.
What it delivers
For whoever owns the store, it comes down to five things.
- Category copy that matches what is in the collection today, including after a season change.
- Internal links between your brand and category pages without anyone placing them by hand.
- The same structure and tone in every language you sell in.
- Customer questions answered on the category page instead of in the chat.
- At PL Line, a 60% reduction in content creation costs.
What it asks of your team
One brief per category and one review round. That is the whole ask.
Everything lands as a draft on a board. You review it per category and approve. Nothing reaches your store without approval. Your SEO agency can hold the same rights and review directly in the platform.
Feedback is applied as a rule in the format document rather than a one-off correction. That makes the second round shorter than the first, and the third shorter than the second.
What it looks like in practice
PL Line, a multi-brand store in Maastricht, carries 82 brands across 259 collections. The briefs came from their own SEO agency, the copy was written per category in four languages, and the agency approved it in the platform.
After the first round the agency sent notes: opening sentences read too much alike and named categories were not linked. Those became rules, and the whole set was written again. The same pattern runs at POSA Studio and NoLabel.
Count it on your own store. How many category texts were written before the last collection change? And how many name a brand without linking to it?