How Ranking Works

Plain-English explanation of how products are ordered on SO-ME — what we use, what we don't use, and why.

The short version:

We rank products by how well they match your query, how popular and current they are, and (where relevant) sustainability and condition. We do not accept payment for placement, and no retailer can buy a higher position by paying us more.

What we use to rank

•Relevance to your query — title, description, brand, category, colour, materials and AI-extracted attributes (occasion, fit, length, etc.) are matched against what you searched for or filtered by.
•Image and content quality — products with multiple high-resolution images, complete descriptions and confirmed sizes rank higher. Listings missing core data are deprioritised.
•Stock and freshness — in-stock items are shown ahead of out-of-stock items. Newly imported listings get a small freshness boost; very old listings without recent updates are softly deprioritised.
•Popularity signals — aggregate click and view patterns at the product and brand level, anonymised. We never use individual user behaviour to rank for other users.
•Condition mix — on category and search pages we deliberately surface a mix of new, pre-loved and rental listings rather than skewing entirely to one. Filters let you narrow if you have a preference.
•Personalisation (when you're signed in) — if you've told us your style preferences, sizes or budget, we use those to filter and re-rank what you see. Personalisation is opt-in and you can clear it at any time.

What we do not use

•Payment for placement. No retailer or brand can pay us to rank higher. There is no "sponsored" placement on SO-ME. Every result you see has earned its position via the criteria above.
•Commission rate. We earn affiliate commission when you buy, and rates vary by retailer. We do not factor this into ranking — a 2% retailer is treated identically to a 10% retailer for the purpose of ordering search results.
•Personal data of other users. Your ranking is based on what you query and (if you're signed in) your preferences. We don't cross-reference your behaviour with other shoppers to nudge what you see.

Promoted edits

Our editorial Edits sometimes feature specific brands or capsule collections. These are always labelled and based on editorial judgement, not paid placement. If we ever run paid editorial in future, it will be clearly labelled as "Sponsored" or "In partnership with" — and never integrated into organic search or category results.

Questions or concerns

Email hello@so-me-app.com. If you think a result is wrongly ranked, mis-categorised or shouldn't be there, tell us — we'll look at it.

What we use to rank

•Relevance to your query — title, description, brand, category, colour, materials and AI-extracted attributes (occasion, fit, length, etc.) are matched against what you searched for or filtered by.

•Image and content quality — products with multiple high-resolution images, complete descriptions and confirmed sizes rank higher. Listings missing core data are deprioritised.

•Stock and freshness — in-stock items are shown ahead of out-of-stock items. Newly imported listings get a small freshness boost; very old listings without recent updates are softly deprioritised.

•Popularity signals — aggregate click and view patterns at the product and brand level, anonymised. We never use individual user behaviour to rank for other users.

•Condition mix — on category and search pages we deliberately surface a mix of new, pre-loved and rental listings rather than skewing entirely to one. Filters let you narrow if you have a preference.

•Personalisation (when you're signed in) — if you've told us your style preferences, sizes or budget, we use those to filter and re-rank what you see. Personalisation is opt-in and you can clear it at any time.

What we do not use

•Payment for placement. No retailer or brand can pay us to rank higher. There is no "sponsored" placement on SO-ME. Every result you see has earned its position via the criteria above.

•Commission rate. We earn affiliate commission when you buy, and rates vary by retailer. We do not factor this into ranking — a 2% retailer is treated identically to a 10% retailer for the purpose of ordering search results.

•Personal data of other users. Your ranking is based on what you query and (if you're signed in) your preferences. We don't cross-reference your behaviour with other shoppers to nudge what you see.

Promoted edits