AI in real estate business
Hi Charu Gulati , as you mentioned market prediction is one thing. However, its like any market prediction. Most of the AI based real estate companies are working in this space only. Also, few startups are working on real estate based customer support by making AI chatbots etc ( Again, chatbot is a standard AI use case and real estate is a use case).
As a landlord and a property hunter myself I feel that AI can become a powerful tool in urban and suburban development. We have already seen an explosion of online services such as Zillow, Realtor, Trulia and more... each providing data aggregation, scoring and trending analysis. The value of these "lenses" as I refer to them is that massive amounts of data is now available to be filter and consumed by multiple end users.
Future applications of AI would logically be based on the data stores created and cultivated by the previously mentioned app owners in conjunction with the MLS owners. Where AI brings the leap forward will be found in the deep analytics of seemingly disconnected factors. For example- choosing a home is based on job proximity for most people, followed by amenities including schools, then it gets complex and seemingly individual. Where AI models can help is in diluting the "uniqueness" into demographic and behavior based decision making. Does a home meeting the size specs sitting in the correct radius from employment and adequate schools meet the unique desires of the potential buyer based on their shopping habits? Odd question at first but consider that a young family, 25 to 30 with 2 kids, tends to shop online and have items delivered versus heading out to shopping centers. Furthermore the delivery routes for primary carriers as well as availability for grocery delivery from local grocery stores is now a factor to this group. Is the home ideal still and worth showing or can it be reduced in scoring in favor of a home that meets all criteria?
As a human working through the nearly incalculable amount of data points it is unlikely that we would be able to divine any pattern, and that is assuming we can approach the transaction without emotion. Sellers, buyers and agents are all emotional beings which throws data analytics for a major curve and often results in poor decisions.
Developers in the real estate business already use different data analytics tools to estimate sprawl and growth on the boundaries of communities. Traffic patterns and retail indicators, new business growth and employment trends are major parts of a decision involving millions of dollars of land improvement over 10 years. Where to build so that housing is available on time in the right area at the right level- think $150k condos versus $750k estates- is critical. I anticipate developers will be first to the AI game as they are most able to bear the upfront cost of developing the tool as well as the most likely to benefit the most from it.
AI can bring buyers and sellers closer together. In the distant future I foresee direct competition to traditional realtors in the form of AI driven virtual realtors with showing agents taking much less of the transaction pie. No longer will there be a flatly accepted 6% transaction fee for the agents. Using AI as the listing, marketing and selling agent the fees can be reduced to bare minimums while improving the access for all.