Collection copy above the grid, below the grid, or split: test it like an operator
Why the fight exists
The shopper arriving from an ad or a menu wants products in the first screen. Every paragraph above the grid pushes the first product down, and on a phone that push is measured in scrolls. The search engine, arriving with no eyes at all, wants text that establishes what the page is about and how it differs from the forty similar collections a crawl away. Both visitors are real, and the layout has to serve them in sequence rather than pick a winner.
The failure modes mirror each other. All copy up top: a page that reads like a brochure and converts like one too. No copy anywhere: a grid of products with prices, indistinguishable to a crawler from every other grid on the store, dependent entirely on links for any ranking it holds. Most stores discover they are running one of these two by accident, inherited from the theme, never chosen.
The split that usually wins
Above the grid, orientation only: what it is, the range logic, one differentiator stated in the store's own terms, price span or material or use case. Sixty words is plenty. Below the grid, the depth: how the range is built, how to choose within it, care or sizing or compatibility substance, and links into the guides that support the purchase. That lower copy is also where the collection's internal linking to sibling collections belongs, in sentences, not in another grid of boxes.
Write the lower copy for the shopper who scrolled past forty products and still has a question, because that is literally who reads it. Comparison detail, the difference between the two constructions you sell, which one suits which use: that is buying information, and it converts the undecided at the exact moment they are deciding. Copy written for a keyword list reads differently and the shopper can tell, even when the crawler cannot.
Test it on two collections before you touch forty
Pick two collections with real traffic and different shapes, one broad, one narrow, and run the split layout against the current one for long enough to cover your normal buying cycle, weekends included. Judge on revenue per session for the collection and organic landing traffic over a longer window, and keep the change if either improves without the other falling apart. A test read after four days is a coin flip with extra steps.
Watch the second order effects while you test: filters used, pages per session from the collection, and the lower copy's own scroll reach if you measure it. If shoppers never reach the lower copy and rankings never needed it, the narrow collection may honestly run lean. The point of testing collections individually is that the right answer is allowed to differ by collection, which theme wide thinking forbids.
Collection copy rots. It names products that left the range, prices that changed, a collaboration that ended in spring. Put collections on the same review rhythm as the products themselves: when the range changes materially, the copy gets read. Stale copy under a fresh grid is worse than none, because it makes promises the grid immediately breaks, in front of the two audiences you were trying to serve.
The copy has to be maintained like stock
Keep one owner for collection copy, with a list of which claims each page makes. When purchasing changes a material or drops a line, that list is how you find every page that mentioned it. Stores discover their copy debt during a complaint. A claims list turns it into a Tuesday afternoon task instead.
Keep reading: Gift guides without a new section: merchandising the catalogue you already have, One offer in the cart drawer: cross selling where the shopper can still say no.
Frequently asked questions
How much copy should sit above a product grid?
Two or three sentences, roughly sixty words: what the collection is, who it serves, and the one thing that distinguishes it. That orients the shopper and gives the page a textual identity without pushing products down the screen. The substantive copy belongs below the grid, where it helps comparison shoppers and still serves search.
Does copy below the grid still help search rankings?
Yes, provided it is real content on the page: range logic, choosing guidance, material and care substance, and links to supporting guides. Search engines read the whole page. What below grid copy cannot do is rescue a page with nothing else going for it, and thin keyword paragraphs written only for crawlers are visible to shoppers who scroll, so write it for them.
What metric should judge a collection layout test?
Revenue per session on the collection, watched alongside organic entrances over a longer window, split by new and returning visitors. Avoid time on page as the decider; it rises with confusion as easily as with interest. Run the test across a full buying cycle, weekends included, before calling it.
How often should collection copy be reviewed?
Whenever the range changes materially, and on a quarterly sweep regardless. Copy naming departed products, old prices, or ended collaborations makes promises the grid breaks immediately. Keeping a claims list per page, with one owner, turns the review into a short task instead of a discovery during a customer complaint.