Artificial intelligence will refuse sellers of illiquid apartments

2025: Introduction of AI predicting the effectiveness of advertising investments

In Gilfond, they developed and began to implement a tool based on artificial intelligence, which predicts the effectiveness of investments in advertising. Thus, the company plans to significantly reduce the cost of exhibiting illiquid objects. The company announced this on January 27, 2026.

The module developed in Gilfond analyzes the object according to 30 different parameters that can affect the prospects for the sale of a particular lot, directly or indirectly. Artificial intelligence analyzes the parameters of the object, compares it with the base and predicts user behavior, finding patterns and predicting the exposure period, as well as the potential volume of investments in its advertising. If an investment in an object's advertising can exceed the planned profit from a transaction, the system warns of such a risk.

As of January 2026, the tool works in semi-automatic mode, we test the capabilities and the percentage of model errors. In the future, we are thinking of fully automating the process, "said Alexander Chernokulsky, director of Zhilfond. - This tool can change the situation established in the market, in which it is difficult for a realtor to predict the sale of an object. For an agent, the logic is simple: the more objects, the higher the chance of a deal, he will receive his commission in any case. The agency can thus receive illiquid objects on the balance sheet, for which the costs of placing ads go, and there are no views and calls. Sometimes we find lots in the database whose advertising costs already exceed the commission we will receive if such a house or apartment is sold in principle. This means that the income from this transaction will not cover either the costs of paying the realtor, or other expenses of the agency: from an office with security and video surveillance, to logistics, accounting, telecommunications, and so on.

According to Alexander Chernokulsky, with the help of the AI module in Zhilfond, they revise the base: brokers find illiquid objects, optimize advertising for them, or completely withdraw from sale in order to prevent future spending on their marketing. In the future, AI is planned to be used to identify risks at the stage of consideration of each object for sale, so as not to take on the sale of deliberately illiquid apartments in principle.

It's more important than ever for real estate agencies to channel funds to where they can turn a profit. Against the background of a high key rate, a decrease in the volume of real estate transactions and tax increases, such a strategy can become decisive, - said Alexander Chernokulsky. - And it is very useful that such a sensitive question today can be entrusted not to a person, whose assessment is often subjective, but to technology, with the help of which you can analyze huge amounts of data, find non-obvious patterns and give out a dense amount of information.