Ranking of offers in the product profile

Offers are ranked in real time when the user loads one of the product profile pages. Machine learning algorithms use a number of factors for ranking, both financial (CPC and the commission amount) and non-financial (buyer interest in product offers, performance in completing orders, and others).

Ranking formula

The position of the product offer (Roffer) is determined using the following formula:

Roffer = ƒ(conv, cbid)

where:

  • conv is the attractiveness of the product offer for buyers, calculated using machine learning algorithms. The calculation takes the following into account: store brand, number of reviews about the store, store rating, completeness of the product description, price of the item, the delivery terms and cost, and other parameters.

  • cbid is the CPC in the product profile that was set by the store. If it has not yet been configured, then the minimum CPC is used.