Ozon content rating: what the score is made of and how to reach the top band

Next to every card in the Ozon seller cabinet sits a number from 0 to 100: the content rating (контент-рейтинг). Sources: seller videos of mid-2026 — one Ozon's own webinar, one with Ozon staff interviews cut in, the rest independent sellers. Two warnings: they contradict each other on how the 100 points split, and we are not papering over that; and Ozon's rules change often, so treat the numbers as a snapshot.

What the score is and what it counts

The Ozon employee's definition: a service showing how much content is filled in on your card. It is split into blocks — media volume, the number of additional attributes that take part in search and filtering, and the other characteristics — summed into a figure in the product list. The scale runs 1 to 100. Ozon's webinar uses it the same way: when you cannot think what else to improve, the rating shows what you left out.

On its effect on ranking there are three answers, and they disagree:

• One seller: with a low rating SEO does not work; the algorithms see an unfilled card and see no point sending it traffic.
• Another: those 100 points are in no Ozon algorithm — though a 100/100 card sells 20% better on average, because filled characteristics widen the set of queries it appears in.
• The Ozon employee: saying we look at the rating when ranking a card in 100% of cases would be wrong; it is one of many parameters.

The shared mechanism is indirect: the attributes it counts feed search and filtering anyway. The split, though, has two irreconcilable versions:

The detailed breakdown: mandatory characteristics earn nothing, since no card exists without them. Search characteristics — those marked with a hashtag icon — give 30 points past 70% of available fields. Additional characteristics give 25 past 50%. The description block caps at 10 without rich content, and only above 500 characters including spaces; with rich content it becomes 20. For media, without 360 photos, one video cover plus three photos and one video suffice.

The second comes from an analysis note in our repository: Media 45, Attributes 30, Text 25; 8 photos plus 1 video for maximum media; a 101-character description floor. We do not hold those transcripts, so it is second-hand and unverified, and the note flags that the source contradicts itself — in one place 101-500 characters is the full 25, in another a 501+ character text without rich content scores 12.5. The gap matters: one scheme puts the weight in attributes (55), the other in media (45). Both lead to the same job — fill in every field you can.

The media block

A video cover (видеообложка) is a clip of roughly 8 seconds playing instead of the main photo, cut from an ordinary video. A detail from Ozon's webinar that most sellers miss: it only plays when a promotion tool is on — trafarety, search promotion, the brand shelf, or reviews for points. Keep it restrained and the first frame informative; not everyone's connection is fast. On conversion, research is quoted for photos from every angle (+35%) and video (up to double).

A sharp counter-example: an advertising-focused seller used to put a cover on every product. On one sitting second in its category, sales stopped for a month; removing the cover brought them back, and the test repeated on several products. He now uses covers almost nowhere, though other sellers tell him they work. What is good for the score is not automatically good for conversion.

A moderation rule from the Ozon employee: no watermarks, no third-party links, no contact details on images — such cards fail moderation and get banned, on creation and update alike.

The characteristics block

The additional attributes the score counts are, by the Ozon employee's description, exactly those that take part in search and filtering — so the block is points and findability at once. The rule from the seller who reached the top 1% of his category: enter everything you know, leave no field empty. Keywords rank very well there, spread evenly across title, description, rich content and characteristics.

Description and rich content

Rich content is the mini-site at the bottom of the card: photo, video and text blocks, built in a visual constructor rather than HTML these days. The commonest mistake is reusing photos and videos already in the gallery; rich content should open the product up with new material, not duplicate the gallery. The analysis note adds that mobile and desktop must be built separately — if only one is prepared it does not appear at all on the other device (we do not hold that transcript). Adding the cover and rich content while creating the card raises the rating.

Technical limits: rich images are 640×640 pixels on mobile, under 10 MB per file; the desktop figure in that video is garbled in the auto subtitles, so we omit it. Discounts, prices and QR codes are forbidden. Do not cram the first slide; one image, one message. Product shots are the backbone — angles, close-ups, open and closed — with model and interior shots complementing rather than replacing them.

Does rich content enter search? Three sources, three answers. The rich design guide says its text fields are analysed by search algorithms, so fill them. Ozon's webinar is asked directly whether rich content is ranked and answers a flat no. The course video says the buyer must expand the description and scroll to the bottom to reach it, so robots index that area last. We are not reconciling the three.

The description takes mid- and low-frequency keywords dissolved into natural sentences: Ozon's algorithms dislike comma-strung keyword lists. Say what problem the product solves, leave no spelling mistakes, and do not copy the Wildberries description over.

Does rich content lift the score? The course video says 10 to 20; the note reports 12.5 to 25 from its second source. But that note also records a live API measurement on 7 cards (2026-07-15) where the text group came back full with a 300-500 character description and no rich content — one measurement in one category, so it neither refutes nor confirms the videos.

Threshold, tracking and the category average

The recommended publishing floor is 80: do not leave a card below it, do not publish without 80+, with bands of 0-39 low, 40-79 medium, 80-100 high. That comes from two videos whose transcripts we lack, so it is unverified first-hand. The sellers we did read aim higher: without 100/100 you have no base and SEO works far worse.

The score sits next to every card in the product list. The note adds that the content rating column on the Products screen lets you download templates for a bulk report, and that under Analytics the products-in-search screen shows each product's rating with its search position against the category average. Aiming above that average beats an absolute 80: competitors fill the same fields.

One boundary: reviews are not part of this score, though they raise conversion by 32% on average, with jump thresholds at 11, 40 and 130-170.

Short summary

• A 0-100 completeness measure, shown per card in the product list.
• Its ranking effect is disputed: SEO dies when it is low; or it is not in the algorithm but a 100/100 card sells ~20% better; or it is one parameter among many.
• Two incompatible splits: 30 search characteristics + 25 additional + 20 description/rich + 25 media; and Media 45 / Attributes 30 / Text 25. We did not reconcile them.
• Media thresholds differ too: cover + 3 photos + 1 video, or 8 photos + 1 video. The cover only plays while a promotion tool is on, and one seller measured it depressing sales for a month.
• Characteristics are the cheapest points: 70% fill for search ones, 50% for additional, doubling as filter and query visibility.
• In rich content: no gallery reuse, mobile and desktop built separately, no prices, discounts or QR codes, mobile images 640×640 under 10 MB.
• Whether rich content counts in search gets three answers; Ozon's webinar says it is not ranked.
• 80 is the recommended floor, the sellers we read aim for 100, and the real benchmark is the category average.