Property: How AI is reinventing our understanding of the sector

What do we know about the property market? In France, more than a million transactions were recorded in 2019. While the pandemic has turned conventional wisdom in its head, it has also confirmed a basic truth – that housing is a central plank of all public policy. Today, Artificial Intelligence offers an unprecedented quality of analysis of the property market.

Artificial Intelligence is a wide-ranging technology used in various ways in our daily lives, from voice recognition on our smartphones to the initial analysis of medical images. For Homiwoo, it offers unprecedented insight into the property market. While its benefits may be clear, people often ask how AI works and how it learns. To answer to those questions, let’s take the analogy of a radiologist’s career path. Starting out, a trainee cannot identify a cancerous tumour. What’s more, learning to do so is not an analytical process. It’s only by looking at images of cancerous tumours, and using the plasticity of the human brain, that the trainee learns how to instinctively recognize them. And it’s the same for AI.

By drawing on a large database, AI identifies the correlations between different pieces of information and can detect current and future trends. Its big advantage is that AI can handle an infinite amount of data, far beyond the capacity of any human being. So, its understanding of the market is more fact-based, rather than being influenced by any cognitive biases that a person might have. For the property business in France, Machine Learning is a unique tool that can handle all the available information, including historical market data, for the whole country. Emerging trends can be identified and quantified in real-time. However, with a technology like this, the issue of data management becomes a central one.

A systemic, all-round approach to the market

For its recommendations to be useful, the machine needs to handle reliable information from a wide variety of sources. It’s a tricky issue, and one that several pioneers in the field have already had to face. To take an example from a different sector, the first facial recognition systems for smartphones were not developed from photos that reflected the full ethnic diversity of the local population. Which meant that some users were unable to unlock their own phones. It’s a similar situation in the property market, ensuring the reliability of data is a key issue. Given the vast quantities of information available today, it’s essential to use Big Data techniques to qualify that data, check its reliability and then process the information, in order to produce a high-quality analysis.

Once those stages have been completed, the benefits of Machine Learning for the property market are numerous. It offers a single scientific approach - a common understanding - for comparing different property assets, whatever the type of transaction. Today, you can compare the sale of an apartment in Paris with that of a house on the outskirts of Lyon. The algorithm will understand the intrinsic differences between the two assets and allow for the different environments, thereby providing a more accurate comparison. In short, Machine Learning offers a comprehensive, systemic approach to the market.

Saving time and rationalising the decision-making process

For the various players in the property sector, AI provides better visibility of the market, particularly in terms of specific financial indicators. It saves time and is able to quantify people’s perceptions of the market. Given the current level of uncertainty, we cannot expect Machine Learning to perform miracles. By its nature, the technology can only handle situations that it has observed in the past, and every crisis has its own challenges and consequences. It can predict the market changes that would follow the building of a high-speed rail connection or a metro station in any given town, but it cannot predict the impact of a second lockdown in the coming months. However, AI has one undeniable advantage: its ability to respond quickly. Since the data is being updated in real-time, AI may not be able to predict the future – but it can certainly help you to prepare for it.

This overall market vision and insight is what we seek at Homiwoo. To achieve that, we have developed a suite of unique solutions to help decision-making, from automated market research to valuation tools. Our team comprises developers and data scientists who are all passionate about their work, and who are harnessing leading-edge technology. We are also pleased and proud to have established a long-term collaboration with the AI Masters programme at France’s École Polytechnique, which has a global reputation for its expertise in the field. Lockdown has only underlined the key importance of housing in our society, making the various aspects of the property market even more crucial. We are delighted to be helping our clients to make the best possible decisions when it comes to choosing a property investment.

William Violet, CEO, and Adrien Bernhardt, CTO of Homiwoo