Optimizing Wildberries Ads: Managing Keyword Clusters, Bids, and Target Positions in ARKNote · digest for October 28
Poster-style vector graphic showing three product card positions in search results with emphasis on the top 5
Note2 sections≈ 3 min

Optimizing Wildberries Ads: Managing Keyword Clusters, Bids, and Target Positions in ARK

Automatic Ad Campaigns (ARK) have officially replaced the old fixed-slot search ads as the primary paid-boosting tool on Wildberries. The mechanics sound simple enough: the algorithm takes your listing's organic rank for a given query and, based on your cost-per-thousand-impressions (CPM) bid, lifts it by a certain number of spots. The paradox is that over 80% of sellers burn through their ARK budgets every day because of two fundamental mistakes: misunderstanding semantic clusters and blindly chasing the #1 search spot.

Let's take off the rose-tinted glasses and look at how the ad auction actually works under the hood.

The Cluster Trap and How to Use Negative Keywords Properly

Wildberries' algorithm does not bid on individual search phrases. It groups dozens of similar queries into a single semantic cluster (for example, "women's evening dress," "long women's evening dresses," and "pretty evening dress" all live in one semantic core). In ARK, you cannot set a separate bid for one phrase inside a cluster—your ads run wholesale across every query in that group.

Here's where beginners make a fatal mistake: they spot one irrelevant query in the report (say, "prom dress" on a strict office dress) and send it to "Exclusions." On Wildberries, excluding a single phrase wipes out the entire cluster, including your highest-converting core keywords.

The iron rule for ARK keywords: never exclude words blindly inside the WB dashboard. First, open an external analytics tool (MPSTATS, MarketPapa, Knopka) and inspect the full cluster breakdown. Only exclude a cluster when more than 70% of the queries inside it are genuine junk for your product.

Chasing Top-1 vs. Holding Positions 5–10 Profitably

The second budget black hole is the illusion that your listing has to sit at #1 to sell well. At #1, CPM is overheated to the limit by big brands and aggressive competitor auto-bidders, spiking to 1,000–1,800 rubles per thousand impressions.

Modern marketplace shoppers don't behave that way—they don't blindly click the very first thumbnail. People scroll the first one or two mobile screens, comparing 5–10 options by price, rating, and cover design. Holding positions 5 to 10 costs two to three times less in CPM (just 350–500 rubles), delivers comparable total click volume, and cuts your cost per order (CPO) by 30–50%.

To keep from selling at a loss, calculate your hard bid ceiling with this formula: Max CPM = Unit Contribution Margin × Click-to-Order Conversion (CR) × Listing CTR × 10. If the live auction rate for your target slot exceeds your calculated Max CPM, running ads turns straight into burning capital. Grounded in how the Wildberries sales funnel and listing conversion work, use external auto-bidders to hold a profitable rank band instead of overpaying in overheated #1 auctions.

Want a deep diveVotes help us choose what to break down in detail. This material has 0 votes.
Sources · 2