Which Products to Feature First in Each Category: Merchandising With Your Real Data

There are two ways to decide which product opens a category: guess, or look at the data. Most stores guess — placing products by gut feeling, or leaving PrestaShop’s generic default criterion in place. Real merchandising is something else: putting forward what the data says will actually sell.

💡

Merchandising: your digital storefront

In a physical store, nobody places products at random: whatever sells fastest goes at eye level. On your ecommerce site, the front row of each category is that shelf. Deciding what goes there — and by what criterion — is one of the cheapest conversion levers around: you’re not touching prices, you’re not touching product pages, just the order.

Fixed rules: the false automation

Sorting by price, newest, or name is convenient because it’s automatic. But these are blind rules: they can’t tell the product that ranks in Google from the one nobody sees, or the one that converts at 4% from the one that never makes it past the cart. A fixed rule applies the same criterion to your star category and to the one nobody visits.

Generic formula

Price, newest, stock, alphabetical. The same for every category, with no idea what your traffic is searching for or what converts. The “automatic” option that’s actually just random.

Your real data

Clicks and rankings from Search Console, behavior and conversions from Analytics, keyword tracking. The order comes from what your store already knows.

What data feeds a good order

  • Search Console — which product pages get clicks and impressions, for which keywords, and at what position.
  • Analytics — which products actually get viewed, added to cart, and bought.
  • Position tracking — if you use Dinorank, its exports add ranking trends and keyword cannibalization to the picture.
  • Business judgment — margin, stock, and season: the data proposes, you decide.

💡 That last piece of data — cannibalization — is the big unknown: two of your own pages fighting over the same keyword. We explain it in keyword cannibalization in online stores.

How Zeyvro Merchandiser does it

Merchandiser brings both halves together: a manual slider that gives you full control, and an order suggestion generated by Gemini from your real data — Search Console, Analytics, and your Dinorank exports if you upload them. The suggestion isn’t a generic formula: it’s calculated from YOUR store’s own numbers. You review it and decide; the module saves the last order per category so there’s always a way back.

Two details built for real-world use: AI responses are cached (24 hours by default), so you don’t burn through API quota every time you open the panel; and you only need your own Gemini key if you use the suggestion — the manual slider works fine without it.

Frequently asked questions

What data does the AI order suggestion use?

Search Console and Analytics, plus position tracking and keyword cannibalization if you upload your Dinorank exports. The suggestion is based on your store’s own data, not a generic formula.

Do I need Dinorank for it to work?

No. Dinorank is an extra data source that fine-tunes the suggestion, not a requirement. Without it, the module works exactly the same.

Does the AI change the order on its own?

No. Gemini proposes an order, and you decide whether to apply it, tweak it with the manual slider, or discard it. And if you apply it and it doesn’t work out, you can go back to the previous order.

Does every AI query use up my API quota?

Gemini’s responses are cached for 24 hours by default, so opening the panel multiple times doesn’t trigger unnecessary repeat calls.

Stop guessing at the order. Let your own data propose it — and you have the final say.

Try Zeyvro Merchandiser free for 7 days →

Zeyvro builds PrestaShop 8 modules from a real store in production. Unobfuscated code.

Enlarged screenshot

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top