Ozon has begun testing its own AI assistant, designed to solve the eternal problem of online shopping — the agonizing choice among thousands of similar products. The pilot project, built on anonymized platform data, is currently available only to a limited circle of users, but it is already clear: this is not just another chatbot, but a strategic move in the battle for the shopper's attention and wallet.

The logic behind integrating conversational AI services into the shopping journey is becoming increasingly obvious. Marketplaces have realized that the only way to retain an indecisive customer is to take on the role of a personal consultant. Instead of tediously sifting through filters and product cards, users can now simply describe their task in words, and the system will offer optimal options on its own.

How Ozon's new algorithm works

The assistant's mechanics are built around dialogue. After switching to AI mode, the user enters a chat where the system asks clarifying questions: about budget, key product characteristics, and the purpose of the purchase. Based on the answers, the algorithm generates a selection, ranking offers by several critically important parameters:

  • Product cost — compliance with the stated budget;
  • Delivery times — speed and convenience of receipt;
  • Seller rating — reliability and reputation;
  • Customer reviews — real user experience.

It is important to note that the system operates on anonymized data: it analyzes search queries, purchase history, a vast array of reviews, and current prices. This approach not only protects personal data but also allows the algorithm to identify objective demand patterns rather than the subjective preferences of a specific user.

Concurrent move by competitors: the new market reality

Notably, Ozon is moving almost in sync with its main competitor. Earlier, at the end of June, the merged company Wildberries & Russ already announced the launch of a similar AI assistant integrated into its mobile app. That service, in addition to selecting products and gifts, can compare similar items and, in the future, is expected to learn to recognize products from photos and take purchase history into account.

This synchronization is no coincidence. Both giants understand that the conversational format fundamentally changes the very logic of search. It is no longer a listing but a consultation, which dramatically increases the likelihood of conversion. A pilot launch with limited access is a standard and sensible approach, allowing them to test recommendation accuracy, infrastructure load, and interface usability before scaling to the entire audience.

My take on the situation: This race of AI assistants is just the tip of the iceberg. In the coming years, we will witness marketplaces transform from storefronts into full-fledged intelligent advisors. The key success factor will not be technological power but the quality of recommendations. Whoever can make selections truly "human" and accurate will gain a decisive advantage in the war for customer loyalty. Keep an eye on developments — this is only the beginning.