The tool is designed to help customers make faster, more informed purchasing decisions by identifying relevant products and recommendations based on user input such as budget, category, or specifications. Available 24/7, iZi reflects a shift in online shopping from a search-based model to a more conversational interface.
Prior to the official launch, iZi was introduced in a soft launch phase and made available to around 50,000 users. Early usage data indicates engagement with the tool, with users asking an average of 2.6 questions per conversation. Around one in three users returned the following week, while a typical session included viewing five to six product pages. According to internal metrics, over 90% of user feedback was positive. The company also reports that approximately 5% of users who interact with iZi complete a purchase directly from the chat interface.
“In the year eMAG marks 25 years, we are launching the first conversational shopping agent developed in Romania. We rely on technology to improve the customer experience and have been investing for more than 10 years in AI and tech infrastructure products. With the launch of iZi, a project we started working on nearly a year ago, we are at the beginning of an important shift, from search-based ecommerce to conversation-based ecommerce, a model inspired by what shopping originally is: a conversation that precedes a transaction. We are proud of the eMAG team that brought iZi to life in less than a year and we will continue to develop new features every month. I invite you to update the app and test iZi,” said Irina Pencea, General Manager eMAG Romania, Bulgaria, and Hungary.
Integrated directly into the eMAG platform, iZi functions as a conversational agent rather than a traditional search engine. It interprets user intent, compares options, and generates tailored recommendations based on three main inputs: the available products on eMAG, the user’s expressed needs, and additional contextual information sourced from the web. When needed, the system also consults external sources such as independent product review websites to support its recommendations.
The assistant has access to the full product catalogue on the platform, processes user queries in real time, and factors in product reviews when suggesting options aligned with the user’s criteria.
So far, iZi appears to be most useful in categories with complex specifications, such as electronics and automotive, where customers often seek guidance similar to that provided by in-store consultants, the company said. Usage is also increasing in categories such as beauty, where product selection often depends on personal needs and additional context beyond standard product descriptions.