Self-purchases live in marketplaces for exactly one reason: ranking algorithms eventually "believe" in conversion. But the key point that you have correctly fixed is that sales depend on conversion, not on the line in the search results. Issuance is a consequence of how the card turns pageviews into purchases, refunds, and repeat orders.
1) “Drawing reviews for new items”
From the point of view of mechanics, it is clear: a new card without social proof converts worse. In practice, the most effective compromise for the market is not to “draw”, but to quickly collect legitimate first reviews through test sales, audience warm—up, bundling, honest answers to questions, and competent instructions in the box. If the reviews are made artificially, there may be no gain: sites are increasingly monitoring anomalies — identical patterns, speed, repetitions, suspicious profiles.
2) “Blocking the negative”
Covering up the negative with self—purchases is often a diagnostic error. Negativity is not a "hindrance", but a signal: quality, expectations, delivery, packaging, size grid, photo. If the cause persists, you will simply accelerate the accumulation of problems: increase orders → increase refunds/claims → the rating drops even more. The "card repair" works more effectively here: correcting the USP, honest photos, instructions, strengthening packaging, working with FAQ, changing the batch/components.
3) Increased ranking by keys: 5-10 buybacks per key and monitoring
It looks logical as a test, but there is a subtlety: the keys cannot be “pumped" in isolation from the unit economy. If the goal is to reduce the DRR and get into high—quality queries, then first the demand is checked with ads for 1-3 days: which queries give additions to the basket, which are empty clicks. Only then does it make sense to strengthen those "key → card" bundles, where the conversion rate is naturally already good. Otherwise, self-purchases become an expensive way to "prove to the algorithm" what the market does not confirm.
4) “Get to the TOP” through the calculation of competitors' revenue
The "crush" strategy only works where there is margin margin and well-established logistics: availability in warehouses, stable delivery, minimal cancellations. Otherwise, you can go up for a couple of days and then roll back: the algorithm sees a spike, but then it fails due to repeated signals (refunds, cancellations, falling conversions). The real consolidation in the TOP is usually achieved not by the number of “buybacks”, but by a combination of: assortment, content, price/value, availability, stable delivery speed, and spot advertising.
To whom is this really (and relatively) actual
- Niches with overheated traffic, where HF/Invoices are expensive and the budget "burns out" quickly — but only if the card is already strong: photo, price, equipment, logistics.
- Categories where sales are almost exclusively based on HF — where any increase in conversion affects visibility.
Who doesn't care — and why
You rightly singled out clothes: the frequency of returns and the variability of sizes/fits make the “conversion boost" less manageable. It is also pointless to "buy" growth.:
- for a product with a bad rating (you accelerate negative statistics),
- for beginners, against brand monopolies (without differences in product and resources, you will lose the economy).
Bottom line: self—purchases are not a growth strategy, but a crutch to maintain the price and the illusion of “higher in output.” If the product is really good, it is often more profitable to invest in content, packaging, an offer and a short advertising test than to pay for artificial signals with the risk of sanctions.
