Artificial Intelligence in Real Estate
July 5, 2018
Artificial intelligence (AI) is a machine’s ability to solve problems by learning over time, similar to the way a brain processes information when making a decision. The term was coined in 1956 by John McCarthy, who also quipped, “As soon as it works, no one calls it AI anymore.”
That has already started to happen. AI technology is now so pervasive that we no longer associate it with science fiction. From protecting you against credit card fraud to recommending shows on Netflix to self-driving cars, this type of intelligence is spreading, as has been made evident by an explosive growth in startup funding since 2010. Investors and innovators alike realize that this technology is the future.
Despite this trend, AI is still in its infancy. Most of today’s computers perform narrow tasks. AI is described as “weak” when all inputs and outputs are defined, like beating humans at chess. This is where real estate websites are today: consumers enter defined inputs (area, max spend, desired number of bedrooms) and the computer produces an output that matches those requirements.
To better replicate the human thought process, computers need to achieve artificial general intelligence, also known as “strong” or “full” AI. This is achieved when computers pass the Turing test, a blind test in which a human would communicate with both another human and a computer to determine which is living and which is not. Human brains are still more powerful than today’s machines because of their architecture and capacity. We possess many layers of interconnected neurons working simultaneously to process information. Artificial neural networks are inspired by this design but still limited by current technology.
So what does the future of real estate technology look like with artificial intelligence? Let’s look at three common tools prevalent in real estate and see how they will evolve with AI.
AUTOMATED VALUATION MODEL
ATTOM Data’s automated valuation model (AVM) is built using regression analysis based on previous home sales. Using property characteristics, we identify similar properties that recently sold to develop an estimate of a property’s market value.
The intelligent system of tomorrow will have the ability to represent and calculate a property’s market value based on many more variables. The graphic above illustrates the complexity of the system, in which each layer represents a single “neuron” within the algorithm. This type of intelligence allows AVM systems to group variables that lead users to an initial approximation that is much more accurate than just looking at home sales trends within an area. If you have the best house on the block, this type of intelligence would take that into account.
It’s feasible that an AVM system could be applied to home buyers, and not just sellers, in the future. What would it look like if you knew the tendencies of someone looking for a home, if you knew their income and what qualities were important to them? You could price homes not just based on the market, but based on individuals.
REAL ESTATE NURTURE SYSTEMS
Many real estate agents and brokerages send out drip marketing campaigns via email to their clients and prospects. These often take the form of new property alerts or even newsletters.
AddressReport is a solution recently acquired by ATTOM Data that sends dynamic personal, targeted content to subscribers based on a property or their area of interest. Because it’s dynamic, subscribers only get notifications when a data point changes — so it always feels new and relevant. Subscribers are more likely to open and engage with email notifications because the messaging — no longer one-size-fits-all — is tailored to their interests.
Personalized nurture campaigns are the future of AI in real estate. Soon you won’t just send email to someone based on their preferences (as is the case with AddressReport), but based on location, time of day, time of year, life event, and even weather. Imagine the power of sending a customized message to someone who you know is about to get married, have a baby, or retire. Industry experts say life event marketing is powerful, and it’s easy to see how it could benefit real estate sales.
There is a trend among real estate portals to start empowering consumers more as they decide where to live. Comparison tools are getting more sophisticated and user-friendly. Mobile apps like RealStir are starting to enable visitors to compare areas much like consumers compare flights on Kayak.com. The real estate industry continues to lag in this area, though. Think of Netflix. In 2012, 75 percent of Netflix’s viewing was based on recommendations generated by intelligence and refined by consumer interaction with the company’s technology — an amazing level of sophistication for two hours of one’s life. Seeing as how a real estate decision impacts how one will spend the next nine years of their life (on average), we can do better.
In the future, real estate portals won’t make recommendations based on physical characteristics (number of rooms, square footage), but based on the consumer’s personal value system. What if a consumer filled out a profile and was matched with properties that best fit his or her ideologies? Using this methodology, portals would deliver fewer but higher-quality recommendations. This is the future and it’s important. Because we’re selling so much more than beds and baths. We owe it to real estate consumers to improve the buying experience.
A version of this article was originally published on www.attomdata.com.