How much is your apartment or house worth? The most classic method is to ask the question to a well-established local real estate agent. Pen and notebook in hand, he will sift through the characteristics of the property (area, floor, exposure, condition, etc.) and will take into account the latest transactions nearby to provide you with an assessment. An “old-fashioned” method that continues to prove itself.
Another solution, many websites such as Bien’ici, Drimki, Efficity, Meilleurs Agents, PAP, SeLoger or Zefir offer online estimates. Artificial intelligence (AI) can enhance the reliability of these algorithms. “We have been using artificial intelligence in our valuation models since 2018, in particular “machine learning”: they are gradually enriched by mixing more and more data on comparable properties, which makes it possible to refine the estimates. »says Thomas Lefebvre, Scientific Director of Meilleurs Agents.
In some geographies, transactions are too rare to provide meaningful comparison data. “Our model will then look for comparable sales in other cities, sometimes located in another region, but for which it has identified similarities, for example on the sociological profile of the inhabitants, the average income, etc. », explains Thomas Lefebvre. This makes it possible to provide a valuation even in the absence of comparable transactions nearby.
Overcome certain criteria
AI can also help buyers find a property. Search engines allowing you to select properties according to the desired characteristics (number of bedrooms, postal code, etc.) have been around for a long time.
“These alerts are not enough. If you are looking for a three-room apartment of at least 65 square meters, the search engines will not offer you properties of 62 square meters which might nevertheless be perfect for you, because they correspond in all respects to your other criteria », says Sarven Ozgullu, co-founder of the start-up Les Clefs de chez moi, which uses AI to optimize the search for properties for its customers.
Its model is thus capable of overcoming certain criteria put forward by the client to offer goods, or even of taking into account criteria rarely offered by traditional search engines. “Our AI is able to read and decipher the implicit in the descriptions of real estate ads to detect certain characteristics that are not indicated in full”, adds Mikaël Couesnon, co-founder of Les Clefs de chez moi.
You have 14.62% .. “”.