From Loss to Profit: How an Ozon Bag Store Reached 934,000 ₽ in Net ProfitArticle · digest for November 18
From Loss to Profit: How an Ozon Bag Store Reached 934,000 ₽ in Net Profit
Article4 sections≈ 7 min

From Loss to Profit: How an Ozon Bag Store Reached 934,000 ₽ in Net Profit

Over a year on Ozon, a women's bag and backpack store raised prices on part of its catalog by 50–60%, walked away from chasing expensive ad reach, and turned a −99,000 ₽ loss into +934,000 ₽ in net profit. Orders grew from 1,078 to 14,000, and order turnover climbed from 3.7 million to 14 million rubles. It sounds backwards: cutting ads and raising prices usually freezes sales. Yet in a case study from Lilia Zharkova's agency presented at MPSTATS Expo 2026, the opposite happened because the team rebuilt the unit economics of the listings instead of pumping in more ad budget.

Why auto-bidding drains budgets on an unrefined listing

The usual product launch goes like this: you throw a listing together in a hurry, stuff the description and attributes with popular category keywords, and flip on auto-ads. It feels like casting a wider net will get your bag in front of more shoppers. In reality, skipping a couple of evenings of keyword cleanup costs far too much.

According to Lilia Zharkova's team, keyword mistakes show up in roughly one out of every three seller accounts they audit. When a listing catches irrelevant high-volume queries, people click through even though your product isn't what they want. They swipe through a few photos, realize it's the wrong item, and close the tab.

That triggers a painful chain reaction. Click-through rate and add-to-cart conversion drop, listing popularity falls, and Ozon's organic composite score slides down with it. To stay visible at all, you have to raise your ad bid, and every click gets expensive. At the start of this project, ad spend ate up 17.4% of order revenue and reached 29.4% per redeemed item, which is how the −99,000 ruble loss piled up.

Before touching any bids, the team cleaned up the foundation. They checked which real queries drove impressions, removed stray words from descriptions and specs, and stopped chasing someone else's high-volume keywords. At the same time, they exported reviews from their own listings and competitors' listings and ran them through an AI model alongside the listing copy. The model quickly surfaced audience segments, buyer pain points, and the exact details shoppers praised or complained about most.

The clearest win came from a unisex backpack. Search queries and reviews showed that male-oriented keywords brought empty clicks with zero purchases and dragged down overall conversion. The team repackaged and repositioned the listing strictly for women, cutting wasted impressions before launching a single campaign.

What to check before turning on auto-ads

If search reports show your listing getting impressions on irrelevant high-volume keywords, auto-bidding will only burn cash faster. Clean the junk out of your product attributes first, study buyer pain points in competitor reviews, and make sure the listing matches what your target shopper actually wants.

How a flagship item and composite score cut bids from 18 ₽ to 12 ₽

Let's quickly cover how Ozon's auction works so the savings make sense. A listing's search rank depends on two parts: the organic composite score (built from sales, conversions, reviews, and rating) and the promotion score (based on your bid and predicted click-through rate).

Once clean keywords help a listing sell steadily, its organic composite score rises. As a result, you can hold the exact same search spot with a lower bid. The bag store proved this with hard numbers: holding second place at a 0.66 composite score initially cost 17–18 rubles per click. After organic performance stabilized, the team held that same second spot for 12 rubles.

To avoid spreading the budget thin across the whole catalog, they built a clear bidding hierarchy:

  1. First, they launched cost-per-click campaigns at an average price with minimum bids to gather baseline data.
  2. They picked one flagship SKU as a traffic driver and gave it a competitive working bid to pull shoppers into the store.
  3. On the remaining bags and backpacks, they set category-minimum bids to collect cheap background traffic.
  4. They separated Search and Recommendations campaigns wherever the working cost per click differed noticeably between the two.

One simple rule helps here: your bid buys your shelf position, while your budget decides how long you stay there and how much demand you capture. Putting a high bid on every SKU in two clicks is easy, but your margin disappears fast. Focusing on one flagship item forces you to watch the funnel closely and trim bids as organic rank grows, sending every ad ruble toward real orders.

Raising prices by 50–60% without crashing conversion

Fixing clicks is only half the battle when rising platform fees and operating costs eat into your margin. If you keep old prices frozen, even strong sales leave less cash in your pocket. In this project, prices on selected items went up by 50–60% over the year.

The fear of raising prices is natural: it feels like shoppers will immediately jump to the listing next to yours. To avoid shocking demand, the team tested price elasticity in small steps while factoring in how the platform's personal customer discount affected the final storefront price.

They raised the price by 300–400 rubles at a time and immediately backed the listing with ads so its search rank wouldn't slip. Then they waited for metrics to settle before measuring purchase conversion. The data showed that a 300–400 ruble step didn't break demand, letting products settle comfortably into a higher price tier with healthy margins.

The tradeoff in this test is clear. During each pause, you spend a bit more on ads to support the listing and wait for numbers to stabilize instead of taking instant savings. In return, the store stops working for empty turnover and keeps real profit on every bag shipped.

Stock levels and holding ad spend at 8%

Cheap clicks and strong margins fall apart overnight if your bestsellers go out of stock. Early on, frequent stockouts were one of this store's biggest headaches. The moment an SKU hit zero, ads paused, organic rank deflated, and after restocking the team had to pay inflated bids all over again to warm the listing back up.

Keeping inventory continuously in stock plugged that leak and eliminated monthly overpayments for repeated relaunches. Steady stock requires tying up more working capital in goods and sticking to a tight replenishment schedule. Over the long run, though, that discipline costs far less than constantly restarting campaigns after another shortage.

What a stockout actually costs you

Hitting zero stock hurts two metrics at once: it wipes out your accumulated organic rank and halts your ad campaigns. Once the warehouse is restocked, the old 12 ₽ CPC won't hold second place anymore, and your budget goes straight back into expensive recovery.

In the end, the work came down to four pillars: clean keywords, steady stock, step-by-step price testing, and lowering bids as organic rank grew. Six months into the rebuild, order volume hit 9.4 million rubles, net profit reached +511,000 rubles, and the ad spend-to-orders ratio settled at 8%. By year-end, the store reached 14,000 orders and 14 million rubles in turnover with +934,000 rubles in net profit, while ad spend stayed around 8% within a 0.6% margin.

Want a deep diveVotes help us choose what to break down in detail. This material has 0 votes.
Sources · 3