We have been selling on Ozon for several years and growing partner stores there. This guide walks through the whole launch: from unit economics to ads and analytics.
Why Ozon, and why unit economics come first
Ozon has more than 500 million site visitors per month. That is a very large pool of potential buyers, backed by mature infrastructure: warehouses, sorting centers, pickup points and its own delivery. Looking at the business without emotion, it is also fairly transparent.
But commissions are high and logistics is expensive, so trading without a worked-out unit economics model makes little sense. Your Ozon business should always start with unit economics and a financial model. Over the years I have seen hundreds of thousands of sellers go bankrupt and close their stores simply because they never ran the numbers.
Why you need to calculate unit economics
Registering a store on Ozon is easy, and there are plenty of instructions for that. The point is that without economics, everything else is meaningless.
You can start with the Ozon calculator. Enter an approximate price, weight, length and width, or find a similar product and load it. The tool then shows the layout for FBO and for FBS.
The Ozon calculator as a quick check
Treat the calculator as a basic check. If it already shows no profitability, a full calculation will most likely confirm that. For example, if a T-shirt at a given commission and logistics cost leaves you a certain net profit, and your purchase price is above that level, there is little sense in entering that product. Taxes, ads, promotions and other charges are not included yet, so real costs will be higher.
If the calculator shows at least some profit, move on to a deeper analysis with a unit economics spreadsheet.
The unit economics spreadsheet
A unit economics spreadsheet collects every cost for your product and calculates what one unit brings you. One unit is one sold item.
Starting budget for three turns
Once unit economics are clear, work out your starting budget. Plan for three turns of stock.
Say you have a budget of $3,300, $5,000 or $8,300. Always set it aside on the basis that you will need to buy the product at least three times. The first batch is ordered and selling while the money is still in circulation. You need to order the second batch. While the second is on its way, you gradually order the third, and cash from the first is only just starting to come in.
To trade on Ozon you need at least $5,000 to $6,700 for three purchases.
Choosing a business form and tax regime
This part depends on your jurisdiction and local rules, so model it with your accountant. What matters for the economics is that the chosen form and tax regime add to your cost base, and that cost must be built into the model in advance.
Also build into the economics at least 10% for advertising and at least 3% for returns. In some categories the return rate is higher: clothing and footwear can run at 20%, 30% or 40% depending on the item. For an average product it is roughly 1% to 3%, sometimes less.
For ads, budget at least 10%. In practice we often bring ad spend down to 2%, 3% or 5%, but at the start a conservative 10% is the right assumption.
Review analysis and customer needs
Next comes niche selection. Collect 20 to 30 days of data, and not from the top sellers. You are not analyzing the best-selling positions but your competitors.
Go into your competitors' reviews and read what people write and which needs remain unmet. Find what competitors' products fail to deliver. You can also dig into product research services. There are plenty of them now, they are useful and many offer free plans.
Analytics services and search queries
Analytics services let you look at queries and products: how many products sell, how many generate revenue, what revenue and turnover look like, and so on. You can also review categories: sales, revenue, number of products selling.
If you or an acquaintance already has an Ozon seller account, analyze search queries in Ozon analytics. Open the search query analytics tab and look at different products and queries.
For example, take dog food. Check weekly query volume, dynamics, add-to-cart, order volume, average market price and so on. You can test a specific product you plan to launch directly through this analytics and see whether there is demand.
This is useful if you already have a registered Ozon store or a friend's store. You will understand upfront whether the product you are considering has money in it and whether there is demand. If there is no demand, there is little sense in entering: earning on it will be hard.
Assessing competition and alternatives
Competition will exist anyway, just as in products with a large market volume. Entering such a product makes little sense; look for options where competition exists but demand is at least present. Then check demand volume: queries, sales, conversions, average order value.
If you have no store of your own or a friend's, analysis through free plans and services is already good and will give you information. After analyzing products you will end up with two, three or five positions that suit you and that you will enjoy selling.
Anchor brands and niche selection
Next, define a market price range, the psychological tiers for your product, and check whether your category has strong anchor brands. An anchor brand is when a query immediately brings one product to the buyer's mind. For razors, for example, people think of Gillette.
Displacing an anchor brand in your category is not impossible, but it is harder. There are categories without an anchor brand that comes to mind instantly. Keep this in mind and prefer niches without an anchor brand, because trading in a category with one is harder.
Then comes purchasing: in China or locally, depending on your approach, economics and judgment. For the first delivery, plan for 30 to 45 days of stock at the expected demand.
Purchasing and store registration
It is better to launch 4 or 5 SKUs of 40 to 80 units each than one SKU of 300 units. At the start you rarely guess which product will take off and bring profit. Take 3, 4 or 5 positions, test more options, see which performs best and then focus on those.
If you are working with a supplier and do not manufacture yourself, start with a test batch: check quality, packaging resistance to impact and moisture, compliance with requirements, protective properties and so on. The test shows whether the product is good and where potential problems lie.
Then comes store registration. The process is simple: fill in the company card and submit the offer agreement for registration.
Product card template and elements
A product card has several elements: title, description, price, images and the cover. The cover should not be the only image. Many sellers add one photo and try to sell that way.
But if you open Ozon today, sellers typically trade with many images. In any category you will see sellers with three, six or more images. It is hard to find a product with a single image now: everyone understands it is not competitive and will not bring results.
What to put on images and covers
So we always make several images. The question is what to place on them.
The first image usually carries the unique selling proposition: what your product is about, plus the product itself. The job of the first cover is to attract attention and stand out through color, a headline or something else.
The second cover is the key feature, the main advantage: why your product is better than competitors'. The third is sizes, usage scheme and so on. The fourth is the package contents. The fifth is usually social proof: certificates, awards. Depending on the product, add everything you have. This builds trust: certificates, guarantees and so on.
Title, attributes and description
The title usually follows a format of 3 to 5 keywords. Keywords are the words potential buyers may search for. You need at least three, ideally five, plus a description of what the product is.
Attributes must be filled in 100%. Attributes are your product's characteristics, and no field should be left empty. Show how your product solves buyers' pains. Not vague claims that your product is the best, but concrete benefits and logic. On marketplaces people read descriptions and product information, and they will not pay for fluff.
Video is a must. Even 15, 20 or 30 seconds is better than nothing. It can be an unboxing: put the product on a table, unpack it, show how it looks. It can be before-and-after, or a demonstration of the product in use.
Pricing and the launch boost
Set a market price. At the start, define the price range in your category, from minimum to maximum. Set a market price or slightly above the market average. To find it, take 20 top competitors for your product and calculate the average.
Why the average, and why slightly above? Because low prices and dumping lead nowhere good in the long run. At the start you can set a low price and try to generate sales, and sales may come. But when you later try to raise the price for more profit, escaping from dumping and low prices is very hard. It can take from several months to half a year. Many sellers cannot escape at all: they raise the price and all positions collapse, with no sales.
Why? Because the Ozon algorithm is used to selling the product at that price. When the seller raises the price to stop dumping, Ozon drops them by 200, 300 or 500 positions, and climbing back is very hard. So low prices and dumping from day one are not worth it.
When you launch a product, you get a launch boost. Ozon gives improved positions, growth in positions and sales. It is giving you sales so you get initial volume and reviews start appearing, so your card begins to grow.
After that boost, focus heavily on growing reviews. Reviews on Ozon are the key tool for high sales. While you have fewer than 50 or 100 reviews, competing with other sellers is hard.
Ways to grow reviews
There are different ways. One is to do nothing: you sell, and reviews come in somehow. But this is unmanaged and very slow.
The second option is reviews for points. You will find this feature in your seller account. Reviews for points are expensive now: we set 100, 150, 200, 300 or 400 points for a buyer to leave a review. You set the rate you are willing to pay and pay when the person leaves a review.
It is expensive: even 100 points is effectively $1.70 out of your unit economics and profitability.
Launching ads and controlling ad cost share
Without reviews it is hard to move forward: neither ads nor organic will work while reviews are few. Once you have at least 20 to 30 reviews, you can launch ads.
After launch, identify products with good starting positions and good ad performance. Look at what works well. You may ask which metric counts as good. There is ad cost share (ACOS), and it should be below 15%.
At the start, aim for ACOS below 10%, or at least 15%. For a launch these are normal numbers, and over time they improve. But from the start, keep it below 15%, and preferably below 10%.
We run all paid ads only on strong products that deliver low ACOS. Products with high ACOS are not promoted and are usually removed. At the start we test cost-per-click: in our tests it gives the most optimal customer acquisition cost.
After launch, focus on ACOS, which you see immediately in ad analytics. Look at the long term to understand a normal ACOS. ACOS of up to 10% of revenue, and lower is better. On our clients' stores it has reached 35% at most, sometimes a bit more. There is always work to bring ACOS down to 2%, 3% or 5%.
Review replies and returns
Next is customer service: replies to reviews and questions. This is mandatory: every review and every question needs an answer.
On returns: they will come anyway. Build a table where you record return reasons: quality, mismatch, sizes and so on. Why? To understand clearly where your weakness is: quality or delivery times.
People write why they return your product. You need to investigate. Then comes weekly analytics: sales volume, net revenue, your ACOS.
Weekly analytics and turnover
Always review analytics for each SKU and each product. Track turnover: how fast the product cycles. Does it sit in the warehouse for 200 days before selling, or does it turn over every month and bring good profitability?
Why look at this? We need to find products with high turnover to cycle our money faster. The faster the turnover, the more we cycle and the more annual return we get.
ABC analysis and assortment management
Also run regular ABC analysis of your assortment. Ideally monthly, at least quarterly. Based on the results, prioritize the top positions, increase their stock in warehouses and scale ads on them.
Category C products are usually removed from sales or sold off. It is important not to allow out-of-stocks. On Ozon this is one of a seller's key sins.
Stock by cluster and AI
Always track turnover for each product by cluster and replenish stock in time so the product does not run out.
Also connect AI tools for analytics, ads and card design. We now produce infographics without designers. We used to pay designers for infographics: a grey cover from a designer cost around $12. The version with Garfield the cat we made with AI. The quality of AI-generated covers is now quite good. AI can also help plan deliveries.
Strategically, I do not recommend betting on Ozon alone.
Summary
Launching on Ozon requires careful preparation: start with unit economics, choose a niche, create quality product cards, grow reviews and launch ads with ACOS control. Do not forget analytics and customer service. And remember: Ozon is a tool for a start, but do not limit yourself to a single sales channel.
