How many metrics really drive ranking
Ozon says around 100 metrics influence ranking and product positions. That is the official line in their articles and instructions. In practice, when we run partner accounts, audits and training, roughly 40 key metrics drive most of the sales and positions. The other 60 or so probably exist, but these 40 carry the weight, and they are easy enough to identify and work with.
The platform never tells you which metrics those are. You find them by testing and checking. Once you do, working with them delivers about 90% of the result. Work with them and you will grow faster than most of your competitors.
The goal is to stop guessing and making chaotic moves. Start using the system's own data so the algorithm pushes your products instead of you constantly adapting to it.
The search visibility tool
Ozon has effectively given sellers an X-ray of the search results. You just need to read it properly. The tool is called "What affects your place in search", or search visibility.
You find it in the Ozon Seller account under Analytics, then Search visibility. There you see the real search results for a specific region and a specific query: who holds which positions, and what their metrics look like.
You also compare your metrics against the top 10 competitors, not just look at your own numbers. And you see where the ranking algorithm deducts points from you and why you are pushed down in ranking and scoring. This is not theory, it is live data.
Open Analytics, then Search visibility, and wait for it to load. A search query loads automatically. Make sure the section that opens is "What affects your place in search". You can read the help section and extra materials there. Then pick one of your products.
Choosing the query and region
Take a product from your range and enter the main query it is promoted for. You can add secondary queries later, but start with the main one. Then choose the region: either where the product is promoted, or where positions are weaker.
For example, sales are strong in one city and you want to reinforce them. Open that city and see that your product sits at position 22. You see your overall score and other metrics right there. Or the opposite: results are much worse in another region, stock is sitting in the warehouse and paid storage is approaching. Open that region and see exactly where you lag and what competitors are doing.
Comparing positions with competitors
Use the same logic for any region. You see the search position, the overall score, and whether promotion is used. You can immediately compare with competitors: their overall scores, promotion statuses, average cost per click, and whether they use pay per order.
An important point: almost all top competitors either bid higher on ads or use pay per order. Most combine both tools, which is why they promote more effectively. Then look at query relevance.
This shows how well you have worked through SEO and keywords, how well they match the query, and how effectively you promote for it. Here you may lag slightly, for example 89 against 93 to 100 for top competitors. Some competitors score lower, but overall that is the picture.
Sales funnel and promotions
The price a customer sees matters: check click-through rate and conversion, meaning how strong your funnel metrics are. Then the boost factors: boosts and participation in promotions. If you use promotions the same way top competitors do, you are fine there. Some sellers skip them, but most use them.
Delivery time of 0 to 1 day puts a product in a good position, though there may still be room to improve. One interesting weak spot is the price index.
Price index and weak spots
In this example there is no boost from the price index, so you miss the 15 to 20 percent that could significantly raise your overall score. As a result you effectively have no advantage there.
To summarize the weak spots:
- Average cost per click is lower than top competitors.
- You do not use pay per order.
- Query relevance, meaning SEO work, is slightly worse than top competitors and needs tuning.
- Overall popularity, meaning conversion metrics, needs work: click-through rate in some places, conversion at some funnel stages in others.
- Promotions are in place and delivery time is fine.
- The price index gives no boost.
Checking why you lag
You need to understand the reason. Perhaps you are compared with a competitor whose product does not really match yours, and that is why you miss the price index boost. Check all these factors and tune them so the overall score rises and product positions improve.
How the algorithm groups data
The ranking algorithm works with a large number of metrics but does not expose all of them. It groups data heavily. Several funnel stages and several conversions are merged into one metric. Popularity, for instance, may combine 5 or 6 metrics into one.
In proper analytics you split them and look at each separately. Even so, the report already gives you a picture of ranking and promotion and shows where you lag behind top competitors. Why do they rank higher and you lower?
One more thing: over the last year, and even the last six months, marketplaces have objectively become more complex.
The match filter and the SEO foundation
Here is how search architecture on Ozon is built. When your product is ranked, there are several steps. The first is the match filter.
It searches for words from the customer's query in the title, description, attributes and so on. In practice, when someone types a query on Ozon, this filter runs first. The algorithm takes your product and your query and searches for a match. If those keywords are missing, you get no impressions for those queries. That means no promotion, less traffic, fewer sales.
That is why correct SEO and well-researched keywords are the foundation of all further promotion, on top of logistics and everything else. Without SEO, all other work gives no result. If your keywords are missing, the product will not be promoted. We build longer descriptions and add more keywords to cover more traffic.
The second step is text proximity.
Text proximity and relevance
Here the algorithm selects the most relevant cards that fit the meaning. The first step is hard filtering on whether the user's keywords appear in your card. The second step is a more careful analysis and a more careful cut.
Sellers often add irrelevant keywords that do not match the query or the product. They do not know how to select keywords and add the wrong ones. At this step the algorithm filters out such clearly irrelevant products.
Base parameters and boosts
The third step analyzes base parameters and calculates a score from 0 to 100. The fourth step assigns boosts to each product. A boost is a multiplier that strengthens your metrics and multiplies the final score.
Your product gets a score during ranking. Some products get 50 points, competitors get 70, others 60 or 40. Then each product receives boosts, an amplifying coefficient.
Think of the ranking algorithm as a formula. The formula adds up metrics and produces a number.
Final score and amplification
The higher that number, the better your positions. But the number is not final. There is a fourth step: boosts that can increase the final number by a coefficient. You may get 50 points, but with a promotion boost you can reach a better final position. An ad boost can also improve your metrics. A price index boost can raise your score too. You need to know which boosts exist and how they raise your final coefficient.
Why a polished card is not in the top
A common question: why do some products with almost no work done rank first, while I polish infographics, text and everything else, and the competitor's card is objectively much worse?
Card quality factors are one thing, but boosts amplify the score further. Your competitor may score lower on card quality, yet with boosts for promotions, ads and price index, the final summed score can be higher than yours with your perfect cards. There is not only the ranking formula but also multipliers you need to keep in mind.
Strategically, these points are not an absolute rating of your card or your product.
Ranking is a race against competitors
The algorithm always compares your products with competitors' products. One competitor has 0 points, the worst score. Another has 100, another 70, another 50. Your product gets 65. Products with worse metrics for the query always rank lower, and products with higher scores rank higher.
Scores change: a living algorithm
Scores change constantly. The ranking algorithm is a moving target. If a competitor improves their card tomorrow, the scale stretches. If they add boosts or something else, your score can drop even if you changed nothing.
Positions drop without changes
Sellers often say: I change nothing, ads worked and still work, I do not touch the images, I ship on time, everything is fine, but positions suddenly started to drop. Did the algorithms break again?
The ranking algorithm is a living thing. Your competitors are not idle either: they take training, study, tinker and change things. One change a competitor implements can make your products drop sharply: metrics fall, ranking indices fall, sales fall.
How to catch up with competitors
Then you sit down again and work out where the competitor moved ahead and why, and close that gap. It is a living process. Some things can be changed in minutes or a day or two, others take months. Let us break that down.
Fast improvements: query and price
The fastest thing to fix is query relevance. Of the metrics we looked at, this is key: rework the title and fill in the attributes. The principle: the more precisely you hit the keywords and the better they are worked through, the higher your final score.
The second is the price the customer sees. This is the price without the Ozon card discount and without the marketplace-funded discount. It instantly affects price indices and the extra boosts you receive. Prices can be played with, and there are strategies to get better promotional boosts at roughly the same margin.
But there is a second category.
Cumulative improvements: reviews and visuals
Some things accumulate over time. First, reviews: the average rating and the total number of reviews count. Why is this important? Because reviews are a ranking factor. The more reviews you have, the better commerce usually is, the better click-through rate, and the better the overall score we looked at.
Second, overall popularity, meaning the conversion funnel of the card: clicks, add-to-carts, favorites, orders. This requires constant work on visuals, infographics, images, titles and extra badges. All of it strengthens the card. Reviews strengthen this popularity too.
Boost factors: promotions, delivery, price index
Now the boost factors that amplify and multiply your base score. The ranking algorithm is a formula that produces a final score, but these factors matter.
First, promotions. Genuinely good promotions worth joining give boosts and higher ranking. Be careful with promotions.
Second, delivery speed. Products delivered in 1 to 2 days always rise in search compared with products that take more than three days. Slower products always rank lower.
Third, the price index. The principle: how attractive your price is compared with other platforms. There are moderate, favorable, super-favorable and unfavorable levels, plus no index at all. This is also a factor.
Advertising as an extra factor
One more factor worth noting is the use of advertising. It also gives points and multiplies your base score.
How to work with the report
The report shows promotion statuses, your points and resulting scores for each tool. Work through it in practice. The principle:
- Analyze your product. First pin it: the report lets you pin your product to the top. It is pinned by default, but if not, find the pin icon and click it.
- Enter your search query, add the city, pin the product, run the search.
- Analyze anomalies: where you lag and which metrics are weak.
- Then work out exactly what to do to improve each metric.
Simple and logical. You just have to do it.
Mistakes when using the tool
Many sellers never use this tool at all. Others use it and make mistakes.
Myth 1: you can just add up all the points to get a score. In reality every parameter has its own weight. 40 points in sales can weigh far more than 40 points in reviews. Working on one metric can strengthen the result far more than working on another. Ideally work on all of them, but some influence the final score more because their weight in the formula is higher.
Myth 2: I improved the description and my points got worse, so it is a bug. Search results are dynamic. While you were editing texts, a competitor may have cut the price, shipped stock to the warehouse, joined a promotion or launched ads. While you made one move, your competitors made their second and third. You are not working in a vacuum but in a competitive market, and the algorithm adapts to everyone.
Myth 3: I have a 25% boost, so I will definitely be first. There is more than one boost, there are several. A 25% boost from promotions is realistic, and you can get more, for example 60%. But it is not the only boost, and others amplify just as much. You can make one boost while a competitor makes three. You make one at 60, they make three at 30 each, and they end up with better metrics. So work across all key metrics and all boosts, and work with this tool and the ranking algorithm continuously.
The position improvement cycle
This should be a continuous cycle of testing and tuning product quality.
- Audit. Collect metrics for all products across all metrics and gather the numbers.
- Base analysis. See where you lag on base metrics: SEO, content, prices, ads.
- Boosts. Analyze where you lag on logistics, promotions, prices and ads, and tune all of it.
- Monitoring. Watch how metrics change. If you still lag, repeat the cycle.
Collect metrics again, find the gaps, tune, monitor. Repeat until the cycle takes you to the top positions. We work with numbers, bring them up to competitors' values, and monitor. If growth is not enough, run the next cycle: review all the metrics again and loop through them again.
The task is simple. You just need to allocate time, do it and have some patience.
Bottom line
Reaching the top is not magic. It is systematic work with analytics and numbers. Understanding how the ranking algorithm works, working with it and influencing key metrics lets you steadily improve positions.
