MELT Investment

How is AI disrupting property investment?

Smart city, building technology, and real estate business. Businessman holding digital tablet with buildings hologram and application programming interface technology













Reviews on Google Maps. Facebook check-ins. And location tags on that Instagram post you just saw. What do all of these have in common? Aside from them being tools we use to socialise online, they are also technologies we use to navigate the modern city offline. 

And a wide array of AI-based technology companies are now using their machine learning algorithms to decipher the unusual data produced by these technologies, and preempt future trends in our cities. More specifically, certain ‘PropTech’ companies are using AI to effectively predict which areas are likely to be a good place for investment, based on these unconventional data points. 

Data complexity in property investment 

Identifying both an area and property to invest in is not a straightforward affair. Although there is a multitude of data available to aid investors in making a decision as to where to focus their efforts, this data is fragmented. The sheer variance of data sources makes it challenging for investors to draw clear hypotheses, and carry out the rigorous due diligence required to make speedy yet accurate investment decisions. 

Within larger market actors, such as institutional investors or REITs (Real Estate Investment Trusts), teams of analysts must sift through millions of data points in an attempt to glean insights from this data. This not only makes it challenging to discern clear patterns but also eats up considerable time. In an industry where speed equals success, many investors find that the best opportunities are simply gone by the time they have deciphered the data. 

So how does AI solve this? 

And that’s exactly where AI comes in. The PropTech sector is where numerous companies are currently applying the power of AI technologies to common real estate problems, such as the one above. AI is essentially defined as the ability of software to firstly analyse and, using machine learning, then interpret data in intelligent and creative ways. Just as a human would, but faster – and with more accuracy. Machine learning is able to effectively recognise patterns in large data sets, and then use those patterns to build an algorithm. That algorithm is then able to make predictions, and these predictions are improved upon the more data points the software ingests and experiences. 

In real estate, AI can be used to analyse large amounts of historical data more efficiently, for example, without the need for human intervention. Given that so many property investment decisions are made on vast amounts of data, and that the process is a cumbersome one that typically slows investors down, it is a sector ripe for AI disruption. But how exactly can AI help investors improve their decision-making? 

AI powers up property investing 

For example, those with even a basic knowledge of property investment understand the importance of proximity to amenities when evaluating assets to invest in. According to Zillow, the price of homes in Boston within ¼ mile of a Starbucks jumped by more than 171% between 1997 and 2014, which is 45 percentage points more than all homes in the city. Similarly, over the past decade, Seattle apartment buildings within a mile of shops like Whole Foods and Trader Joe’s appreciated in value faster than others.

An AI algorithm can combine these data points with newer, more unconventional ones, to build a granular picture of an area, right down to the street level. New data sources, such as reviews of local restaurants, social media check-ins, and mobile phone signal patterns are becoming more and more relevant. Even more importantly, these trends can be seen with such granularity that it is possible to make predictions about two different buildings in the same postcode. 

WeWork is a stellar example of the above in action. By teaming up with a giant yet little-known big data company called Factual, WeWork has created a very specific index of amenities which the neighborhood of any office must-have. This index contains everything from specific brands of coffee shop right down to proximity to good fitness facilities. Thanks to AI, WeWork can easily identify locations that fit their investment criteria with the click of a button. The strategy has certainly paid off, with WeWork adding 7.7 million sqft of office space to their business in 2019. 

AI applications outside of the investing arena  

And it doesn’t stop there. AI is also revolutionising the process of buying property at the individual level, such as when buying a home. Startups such as Proportunity not only hope to make the process of becoming a homeowner easier but also more affordable. 

In a nutshell, Proportunity aims to identify the parts of London which, according to certain markers, are on the ‘cusp of gentrification’. They market their product specifically to the millennial demographic, who are often on middle incomes, yet who struggle to get on the property ladder. 

Since Proportunity can predict the future price increases a certain property is likely to experience, they are able to offer a loan against the anticipated value of an asset, as opposed to today’s lower value. This is, without a doubt, a win for many aspiring homeowners in the capital. 

What’s most interesting, however, is the data that companies such as Proportunity use to effectively recognise gentrification. Whilst certain data points such as crime rates, unemployment levels, connectivity, and school ratings are fairly conventional, others are more unusual. The rise of social media, for example, allows us to now track consumption habits according to where people check-in. 

And, perhaps most peculiarly, Proportunity examines data on chemical compounds found in local sewers in order to determine which drugs are used in certain areas. A decrease in crack cocaine accompanied by an increase in cocaine deposits typically marks a changing social demographic, according to the company.