Buyouts and Reviews on Marketplaces: A Full Methodology for Safe Purchases and Rating Work

Over the past five years the term "buyout" has run a full cycle: from a secret technique among top sellers to a crude service selling "50 buyouts for 5,000 ₽ with a basement full of accounts from the Sadovod market". In 2026 it is a demanding discipline again — because marketplace algorithms have learned to tell a real buyer from a bot farm across dozens of signals. Half the market still works off 2020 playbooks and collects fines. The other half believes buyouts are about "tricking the algorithm".

Both positions miss. A buyout is a purchase made in order to give the algorithm a demand signal for a specific keyword. Nobody is tricking the algorithm — it is being shown that the product can sell for this query. From there its job is to widen reach, watch how the product card behaves on organic traffic, and lock the position in if the conversions are genuine.

This guide is a methodological breakdown. What a buyout is, when it works, how to calculate pace, what separates a white-hat model from a bot farm, how WB, Ozon, and Yandex Market differ, and how reviews weave into the strategy. Within cluster C of our semantic core (~25 keywords on buyouts plus 16 on reviews) this material is the Pillar. Platform-specific guides are published separately. The guide on marketplace promotion is the parent one and is best read first: it explains the general logic of visibility and the 8-module MPO system. This one goes deeper into the Boost and Trust modules.

This is the most underrated law of marketplaces. Most buyout services work by SKU code or direct link: the buyer receives the SKU, opens the product card directly, and places the order. From the algorithm's point of view this is a purchase that arrived "from somewhere outside" — it counts toward the overall weight of the product but is not attached to any search query.

The result is that buyouts only move the "overall weight" of the card: rankings rise a little for 1–3 days, then roll back. For specific keywords nothing gets locked in. The algorithm has no idea which queries the product should be shown for more often.

A proper buyout looks different:

1. 1The buyer opens the marketplace app or website like an ordinary user.
2. 2Types a specific keyword into the search bar — "moisturizing face serum", "non-slip yoga mat", "TV wall mount 32 inch".
3. 3Scrolls the results. May compare several products. Sometimes opens a competing item, reads its reviews, and comes back.
4. 4Clicks our product card. Studies it. Adds it to the cart or to favorites.
5. 5Places the order. Pays with their own card. Collects it at their own pickup point. Sometimes leaves a review 5–10 days later.

From the algorithm's point of view this is proven demand for a specific keyword. The purchase counts toward revenue for that query, lifts rankings, and — given enough volume — the card locks into the results organically.

A simple corollary of this law: a brand or color filter outweighs a purchase made on someone else's keyword. If you want to lift "face serum" but the filters carry the brand "LABUBO", the purchase goes to the query "face serum LABUBO" and does nothing for the parent cluster. This is a common and expensive mistake. It is covered in detail in the Buyouts 2.0 guide.

This is not one universal tool. Different jobs call for different scenarios, budgets, and pacing. Eight scenarios where launching buyouts makes sense:

1. Lift rankings on specific keywords. The card sits at position 60 or below and nobody scrolls that far. The goal is to pull it into the top 10–25 on priority queries. Buyouts move from LF to MF keywords, then to HF.

2. Launch a new product. The algorithm does not trust a new card. Starting signals let it out of the sandbox: 10–30 purchases across priority LF clusters in the first two weeks.

3. Shake up a stalled card. Rankings are steady but not growing. An impulse on new clusters you have not tried yet sometimes breaks the plateau.

4. Prepare a product for its season. Four to six weeks before peak demand, claim the shelf and lock in. In winter you warm up "heaters", in August "school backpacks", in November "New Year decorations".

5. Recover a card after an out-of-stock. A month with no stock rolls rankings back by dozens of rows. Warming up through buyouts restores visibility faster than advertising alone.

6. Lower CPM on expensive clusters. If a keyword carries a high bid but organic is close (positions 10–25), targeted buyouts pull the card into the paid slots through organic. Advertising gets cheaper automatically.

7. Collect starter reviews. A real purchase means the option to leave a review. The buyout-plus-review combination works better than a pure "buy a review" service.

8. Give the algorithm an evidence base. Sometimes a product sells but the algorithm does not believe in its clusters — it cannot tell which query to attach it to. Buyouts across different keywords help distribute the weight.

Goals, budget, and pace differ in every scenario. Running the same "50 buyouts" against all eight is pointless.

The market has three main formats:

That price is the cost of the contractor's service, with no product cost, marketplace commission, taxes, or delivery included. The full cost of a single purchase is made up of:

• the buyout service fee (325–375 ₽)
• the cost of the product itself (from its shelf price)
• the marketplace commission (15–25% depending on category)
• delivery and returns where needed
• taxes, depending on your tax regime

The simple formula is full cost of one purchase = service + commission + product price. The service fee is only part of the budget. What eats a project hardest is the cost of the goods themselves: at a 500 ₽ price tag and a 325 ₽ service fee, one purchase runs roughly 900–1,000 ₽ all in. On a 300-buyout project that is around 300,000 ₽.

Free entry points that let you start with no spend: a free consultation, a free product card review, keyword selection, a mini SKU audit, and a growth map for a product line.

Most of the market runs buyouts as a ladder: 2-4-8-16-32 a day. The idea is to pump activity fast, watch rankings peak, and close the project there. The algorithm spots it instantly: conversions jump for a week, and the moment buyouts stop they collapse to zero. That is an unmistakable "the spike was fake" signal.

The right pace is a plateau: smooth growth, a hold, a smooth decline. Alongside it, organic traffic builds through the second month.

The logic of the plateau:

• Days 1–3. Start at a low pace. The algorithm sees the first purchases on the keyword. The low-traffic card test kicks in.
• Days 4–10. The pace rises smoothly. Purchases come in steadily. The algorithm starts widening reach.
• Days 11–20. A plateau at optimal intensity. You watch CTR and intermediate conversions — if they hold, the marketplace starts handing over organic traffic.
• Days 21–30. A smooth decline. Part of the buyouts is withdrawn because organic has already caught on. The card keeps growing without support.

By the end of the month the project looks like this: the total purchase count is not much higher than with a ladder (often it is lower), but the ranking chart is completely different. The ladder gives a peak and a cliff; the plateau gives a stable hold.

That is the answer to the popular question "what if we just do fewer buyouts?". Fewer does not work if the pace makes no sense. A sensible pace at the same volume gets locked in by organic traffic.

There is a myth that marketplaces "check every purchase". That is technically impossible: Wildberries processes tens of millions of transactions a day, and storing and analyzing every session would cost more than the fines are worth. Marketplaces work differently: they compare the totality of signals and quickly filter out suspicious patterns.

In 2026 the algorithms track at least nine markers:

Briefly, one by one:

1. 1Payment logic. Payment method, issuing bank, transaction pattern over time. Virtual cards from a single BIN plus identical payment histories are an instant flag.
2. 2Device and fingerprint. IP, screen resolution, font set, browser version, installed certificates. An antidetect browser can be tuned, but the same tuning across 100 accounts is recognizable.
3. 3Geography. The IP region has to match the pickup point region. Orders concentrated at one or two pickup points is the top flag of 2025–2026.
4. 4Account age. An account created the day before the purchase that immediately ordered exactly your product is suspicious. An aged buyer account with a long life cycle buys varied things, regularly, across different categories.
5. 5Purchase history. How many SKUs the account bought before and which ones, in what categories, at what frequency. An account with a single purchase is a flag.
6. 6Session pattern. A real buyer scrolls the results, compares alternatives, reads reviews, sometimes comes back an hour later. A "opened it, straight into the cart, paid" journey in 100% of cases is unrealistic.
7. 7Pickup rate. Real buyers decline orders at the pickup point — in our data, in 5–15% of cases. An account that collects 100% of everything it orders is a flag.
8. 8Review rate. If a category normally sees one review per 100 purchases and you get 50 reviews in a week, that is a signal. A disproportionate wave is the main detector of bought reviews.
9. 9Intermediate conversions. External traffic (bloggers, direct links): 100 visits to 80 add-to-carts to 55 orders to 50 pickups. Organic: 100 to 15 to 10 to 5. A 10x difference. If buyouts run through external links, the conversions land outside organic ranges, and that shows up instantly.

That is the answer to the popular question "what if we just do buyouts carefully?". Careful does not mean "one a day from different cards". Careful means no red flags across 3 or more markers at once. One or two can be forgiven. Three means a review.

Most cheap buyout services run the same setup: one or two antidetect browsers, virtual cards from a single BIN, throwaway accounts, and everything concentrated on 2–3 pickup points in Moscow. The effectiveness of such services in 2026 is 30–50%. Half the buyouts fail, half the reviews fall off, and fines land periodically.

An aged buyer account has a long history: 6+ months of activity, regular purchases across different categories, unique payment logic (a real card, not a virtual one), a realistic fingerprint, an IP in the right region, and a pickup point in that same region. Such a persona behaves like an ordinary buyer — because it is an ordinary buyer who occasionally carries out our tasks.

This is a fundamental difference, and it explains why a cheap service usually turns out more expensive:

• A "100 buyouts for 30,000 ₽" service with a 50% success rate delivers 50 successful purchases. Effective cost per purchase: 600 ₽.
• A "100 buyouts for 50,000 ₽" service with a 95% success rate delivers 95. Effective cost: about 526 ₽.

And that is before fines, ranking rollbacks, and lost time.

The most common planning mistake is answering this in the abstract: "will 50 a day lift it?". It depends on the niche, the competitor, the keyword, the search volume, and the product price.

A simple way to estimate is through the revenue of a top-10 competitor on the same keyword. If a product in 5th to 10th place in the category makes 500 purchases a month on that keyword, then to stand there you need a comparable volume on that keyword.

Rough benchmarks:

• Low-frequency keyword (LF). 5–15 purchases over 2 weeks is usually enough.
• Mid-frequency (MF). 30–80 purchases.
• High-frequency (HF). 100–300+ purchases, sometimes more.

Why start with LF, then MF, then HF rather than going straight to HF:

• LF keywords are niche, small, and high-converting. It takes fewer purchases to take the top.
• The algorithm first tests a card on safe, low-traffic clusters — that is its natural logic.
• If conversions on LF keywords are good, the marketplace scales the card up to MF, then to HF.
• Go straight for HF and the marketplace does not believe you and will not widen the reach.

This works on any marketplace. The principle is universal.

A buyout strategy does not transfer directly between platforms. Briefly, one by one:

Buyouts on Wildberries

The largest platform in Russia, the strictest anti-fraud among the classic marketplaces, and the harshest dependence on logistics and warehousing. What to account for:

• Warehouses decide everything. No stock at the nearest warehouse means lower visibility, and buyouts will not fix it. Ship to FBO first, pump later.
• CTR in search results is critical. If it is below the category level, buyouts will lift the position for a day and let it fall back. Fix the cover image and infographics first.
• A mobile audience. WB lives on mobile. Scenarios are modeled around mobile behavior.
• Anti-fraud is strict but does not pick on the white-hat model. Aged buyer accounts pass, throwaway accounts do not.
• Time to effect. First signals at 14 days, lock-in at 60–90.

Queries like "buyouts for wildberries", "product buyout on wb", and "order buyouts on wildberries" all point to this scenario. The deep breakdown of WB mechanics specifically is in the Hub guide Buyouts on Wildberries: 3 buyout types, the anatomy of a wave from day 0 to day 30, 7 safety markers, and how it ties into reviews.

Buyouts on Ozon

The second platform by turnover, and it differs from WB in three ways:

• Recommendation shelves are a separate channel. "Similar products", "frequently bought together", and "customers also buy" work independently of ordinary search. They have their own buyout type: the shelf buyout.
• Promo codes and rFBS draw suspicion. When the marketplace collects less than its full commission (rFBS) or the product is bought at a discount (promo code), Ozon pays that purchase more attention. Buyouts with promo codes are a common trap. Full price on FBO is better.
• The desktop audience matters more. In expensive categories a meaningful share of buyers is on desktop. Scenarios are modeled for both mobile and desktop.

Queries like "ozon buyouts" and "order ozon buyouts" are about these specifics. Without understanding recommendation shelves, buyouts on Ozon deliver 30–50% of their potential. The deep breakdown of Ozon mechanics is in the Hub guide Buyouts on Ozon: 3 buyout types, pairing with advertising to cut ad spend share (DRR), and 4 safety rules (promo codes, rFBS, desktop, promotions).

Buyouts on Yandex Market

The third platform, usually last on the priority list. Its specifics:

• Seller rating. On Market one of the key signals is the seller rating, not just the card. So buyouts here are often tied to the goal of lifting the seller rating across an SKU line rather than for one product alone.
• Product card and seller are separate. On Market a single product can be sold by several sellers within one card. The buyout strategy accounts for how you win share of impressions away from the other sellers.
• Integration with Yandex search. Part of the traffic arrives through general Yandex search, which sets Market sharply apart from WB and Ozon.
• Less competition in most categories. The bar for entering the top 10 is lower. Sometimes 50 buyouts on Market give the same effect as 300 on WB.

Queries like "yandex market buyouts" and "yandex buyouts" are about this.

If nobody sees the card, nobody reads the reviews. But once visibility catches on, reviews become the strongest conversion factor on marketplaces. A good photo review showing the real product in real conditions lifts add-to-cart conversion by 10–25%.

Accordingly, cluster C of our semantic core covers not only buyout queries but also the "buy reviews wb / ozon / yandex market" group. It is part of the same discipline, and working on reviews separately from buyouts is almost always a mistake.

4 Review Types and Where Each One Fits

That is the service price. The full economics look different.

Effective Cost per Review: Why Cheap Usually Ends Up Expensive

The real price of a review is not the one on the price list. A review has to be paid for in full: the service fee plus the buyout itself, the product, the platform commission, and taxes. On top of that a share of attempts falls away — the buyer did not leave a review, the review failed moderation, or the review was deleted a month later.

What you have to calculate is the full cost of an attempt divided by the share of reviews that actually stayed on the card. Take your own numbers from the contractor's reports: they vary several times over between providers, and they are exactly what turns a cheap price list into an expensive result.

The detailed calculation and formula are in the Safe Buyouts guide.

Review Pace Follows the Same Laws as Buyouts

The main rule: reviews must not break the proportions of the category. If a niche normally sees one review per 100 purchases, fifty reviews per 100 purchases look unnatural. The marketplace will strip half of them and the card picks up a flag.

A healthy proportion is 10–75% reviews out of all buyouts in the first month, after which the pace levels out to the category average.

What else matters:

• Do not leave a review right after delivery. Real buyers use the product for a week or two and then write. A 5–10 day lag is normal.
• Unique texts, not templates. If 30 reviews all open with "Great product, thank you!", that is spotted in a minute.
• Do not push ChatGPT reviews through blind. Language models generate similar constructions and marketplaces have learned to filter them. A human editor on top is basic hygiene.
• Spread the formats. Purely text reviews look unnatural. There should be photos and at least a couple of videos.

One of the most popular queries is "delete a review on wb" or "remove a negative review on ozon". The direct answer: a contractor cannot delete a review. Only marketplace moderation deletes reviews, and only on clear grounds (abuse, advertising, spam, a photo attached by mistake).

What does work is burying negatives: fresh, realistic reviews push the old ones onto the second and third page. The buyer sees the first page, and it is positive. Conversion recovers and the rating climbs.

How it works:

1. 1Audit: how much negativity there is, of what kind, and which pages it currently sits on.
2. 2A hypothesis about the causes: packaging, batch, description, or expectations.
3. 3In parallel, a recommendation to the seller (if the negativity is about real problems, fix the source).
4. 4A plan: 10–20 fresh, realistic reviews (text plus photos) spread over 2–4 weeks.
5. 5The effect by the end of the month: rating 4.2 to 4.6, negatives on the second page, conversion recovering.

The hard truth is that most of the market runs on a "money up front, we will sort it out later" model. The defense is simple, and it is ignored in 80% of deals:

1. 1A contract is mandatory. Not "we agreed in chat" but a properly signed document covering scope, volume, deadlines, and reporting. Without it you have no ground to stand on.
2. 2Purchase receipts are mandatory. No receipt means the work was not done. Full stop. This is the basic artifact a contractor must supply for every purchase.
3. 3shipment IDs (SRIDs) and review IDs are mandatory. Every review on a marketplace has a unique identifier. No ID, no proof.
4. 4Transparent reporting on every wave. Not "we did 100 of them" but a structured table with dates, keywords, pickup points, accounts (anonymized), receipts, and reviews.
5. 5Damage control. Some goods are ruined in transit — dented, broken, or swapped for something else; across our projects this runs to a few percent of volume. On buyouts it means the contractor receives the product in the wrong condition. Up to 5% is normal. Above that you need to work out who is at fault. The contractor should compensate what was lost at cost plus delivery.

A story from practice: "I ran buyouts with a budget of 1.5m ₽ on goods and 600,000 ₽ on services. The reviews did not pass, a fine came in, and we stopped the buyouts. 30% of the goods were 'lost'. I want my 300,000 ₽ back, and they tell me the service was rendered and they know nothing about it." This is a typical situation. With transparent reporting and a contract it would not have happened.

Before starting any buyout campaign, walk through this list:

• [ ] The campaign goal is stated. Not "lift the card" but "put SKU X in the top 15 on keyword Y within 30 days".
• [ ] Keywords are collected. Not off the top of your head but through analysis of real demand. See the promotion Pillar and Buyouts 2.0.
• [ ] SEO is updated. The keywords you will be buying on appear in the card title and description.
• [ ] Stock is at FBO or the nearest warehouses. With no stock, buyouts are pointless.
• [ ] Cover image and infographics are checked. If CTR in search results is low, buyouts will give only a short-term effect.
• [ ] Budget is calculated at full cost. Not just the service but the product plus commission plus taxes.
• [ ] The pace is a plateau, not a ladder.
• [ ] The review plan is folded into the overall plan: format, pace, texts.
• [ ] The contractor has aged buyer accounts, a contract, and receipts. Not a bot farm.
• [ ] Metrics are on a dashboard. Rankings on priority keywords, CTR, conversions, and DRR, updated daily.

"Is this even safe? WB bans people for buyouts." WB bans crude schemes: throwaway accounts, everything concentrated on 2 pickup points, template scenarios, shock waves of conversions. Aged buyer accounts with a long life cycle, realistic scenarios, and a pyramid pace get through. In 7 years of work we have had no systemic fines on client accounts.

"How many buyouts do I need?" It depends on the niche, the competitor, the keyword volume, and your current position. There is no universal number. Roughly: 5–15 for a low-frequency keyword, 30–80 for mid-frequency, 100 and up for high-frequency. The exact figure comes from analyzing top-10 competitors.

"How long until it works?" First signals at 14–30 days. Lock-in at 60–90. If someone promises "top 10 in a week", that is a ladder with a rollback, not lock-in.

"Can any product card be lifted?" No. If the product is weak on price, rating, or quality, buyouts will not magically make it sell. That is obvious at the intake consultation and we will say so honestly.

"Can I do buyouts only, without SEO and reviews?" Technically yes. But without proper SEO the buyouts miss the keywords, and without reviews the card is visible but does not convert. At minimum we recommend keyword selection before you start.

"Can I buy reviews on Wildberries separately?" Yes, reviews are a standalone service. But they usually deliver more paired with buyouts. Pure review buying without buyouts is either burying negatives or building a starter pool for a new card.

"Is buying reviews on Ozon legal?" It is a gray area, same as buyouts. The marketplace does not endorse it but does not actively ban it either, as long as the review follows a real purchase, is written uniquely, and stays within category proportions. Template waves from a bot farm are a flat no.

"What if we already got a fine?" The situation needs an audit. Sometimes the strategy can be retuned and continued. Sometimes it is better to let the card cool down for 2–4 weeks before the next wave.

"What about rFBS and eDBS?" When the marketplace collects less than its full economics, buyouts draw more suspicion. Full price on FBO is better. If you work only on rFBS, the strategy is agreed individually.

"Can we do it through bloggers instead of buyouts?" Bloggers are external traffic. It has different intermediate conversions (100 to 80 to 55 to 50), which the algorithm reads as atypical. The effect on rankings for search keywords is limited.

"What about returns?" Realistic scenarios sometimes include returns — that is part of the credibility. 5–10% returns across scenarios is normal and even useful. Zero returns is a flag.

"What if I just use friends?" If your friends genuinely live in different regions, have different cards and different pickup points, and really did search for the product, that is exactly what proper buyouts are, only at a tiny scale. The problem is that you do not have 300 friends.

"Which marketplace should I start with?" If you are already on WB, start there — it has the most traffic. If you are planning a launch and can choose, a niche product is sometimes better off starting on Yandex Market, where competition is softer.

"What does a buyout cost with the service and the product on a 300-buyout project?" A rough benchmark is 250,000–400,000 ₽ depending on the product price. The service is 100,000–110,000 ₽, the rest is product, commission, and taxes. The exact figure comes out of a consultation.

"Is this suitable for large brands?" Yes. Trusty is deliberately positioned for enterprise: aged buyer accounts, project reporting with receipts and SRIDs, contracts, and financial acceptance documents. It stands up to finance and compliance departments.

"And for agencies?" We often work as a white-label subcontractor. The agency owns the client relationship, we handle delivery. Transparent reports, receipts, contracts.

A few cases where we honestly say "not now" or "not us":

• A card with 100% visibility on the keywords that matter. There is nowhere left to grow — buyouts would pump what is already maximized.
• Systematic problems with the product. Defects, a bad batch, constant returns — fix the source first, promote afterwards.
• "Do 30 buyouts and forget about it." That is not our model. Under 200 buyouts per project does not work economically.
• Turnover below 2m ₽ a month. The full cost of a purchase plus service plus product plus taxes will eat the margin faster than we can build momentum.
• Fully regulated categories — medicine, supplements, weapons. Discussed case by case; sometimes we take it on, sometimes we decline honestly.
• Market as a side channel, not a priority. If 80% of turnover is on WB and Market gives 5%, spreading yourself thin makes no sense.

• Marketplace Promotion: A Systematic Guide — the parent Pillar on MPO. Explains how buyouts fit into the 8-module system.
• Micro-Differences, Macro Results — how visibility is built and how behavioral signals work. The foundation without which nothing else holds.
• Buyouts 2.0 — strategies by goal (ranking growth, reviews, lower DRR, launch, out-of-stock, season), the plateau pattern in detail, and budget calculation through competitors.
• Safe Buyouts — marketplace markers, aged buyer accounts, effective cost per review, and protection against dishonest contractors. The full effective-cost formula is in the code.

If the task is still taking shape, take the free product card review. We will show your share of search results across priority clusters, find the gap keywords, and tell you whether the goal is realistic. Roughly a third of these reviews end with "you do not need this right now" — sometimes redoing the cover image or the packaging pays off more than pumping visibility.

If the strategy is clear and you need someone to run it, look at the Trusty buyout service and write in for a consultation. Up front we ask about turnover, niches, current DRR, and the format of cooperation (brand, agency, or seller). What you get back is a 90-day plan with a budget and priorities.