AI in Real Estate

How AI Has Been Shaping Real Estate For Years

Artificial Intelligence is increasingly shaping many aspects of our world and the real estate industry is no exception.

And while much of the news lately has been on game-changing chatbots like ChatGPT, the truth is that AI has been transforming the way we buy, sell, and manage properties for years.

In this post, we will explore some of the ways that AI has been used in real estate and take a closer look at how those uses have impacted, and continue to impact, the industry.

But first, let’s be clear about our terms…

What Is Artificial Intelligence?

The term “artificial intelligence”, or AI, usually refers to computer programs that perform tasks associated with human intelligence, such as perception, learning, problem-solving, and decision-making.

While some people associate AI with complex systems like self-driving cars, facial recognition, or even technologies from science-fiction, the term can also include simpler algorithms performing much more basic tasks, either by themselves or in concert with others.

In short, the complexity of an algorithm does not determine whether something counts as AI or not. Instead, it is whether the task is one that typically requires human intelligence, even if the AI does it in a very non-human-like way.

Does that all sound a bit vague? It is! But that’s because there are still ongoing disagreements about what exactly is or isn’t AI.

So why are we bringing it up? 

Because in order to appreciate the many ways that AI is already shaping our lives, we first need to broaden our frame of reference, seeing beyond the examples that are just making today's headlines.

What Is Real Estate?

Real estate is a field or industry that encompasses a wide range of activities related to physical land, buildings, and structures, as well as the legal rights and interests associated with them.

Now you might think we're being really pedantic. As if you don't know what real estate is! But again, it's important to be clear here, because AI has had some very different implementations and impacts in different parts of the real estate world.

Which parts are those? Well, we can divide real estate into two broad but related categories of activity:

• Real estate may refer to property investment and management, including activities such as purchasing, developing, and leasing property.
• Real estate may refer to the work that real estate agents and brokerages do to help buyers and sellers navigate the process of buying or selling a property, including marketing, negotiation, and paperwork.

So let's look at some of the ways that AI has been used in each of these spheres of real estate...

Uses of AI in Real Estate Investment & Property Management

Successful property investment and management requires a delicate balance of analysis, forecasting, and risk management.

Previously, this required a combination of experience, knowledge, and sound judgment. But real estate investors and managers have increasingly been relying on AI to incorporate and assess large amounts of data and to produce actionable insights.

For example, these AI systems are being used to identify:

• Inefficiencies in property usage
• Optimal rents
• Properties with appreciation potential
• Properties that owners may be likely to sell

By identifying hidden trends and predicting future conditions, AI has allowed property investors and managers to make informed decisions about where to invest and how to manage their portfolios. Some significant players in this space are:

• SkylineAI - Offering AI assistance for commercial real estate investment
• HouseCanary - Developing algorithms to analyze and forecast residential real estate trends
• Veros - Using AI for property valuations and risk management

AI is also increasingly being used to directly manage properties, optimizing things like security, space and energy usage. Companies in this space include: Verdigris, BuildingIQ, and Pointgrab. 

Then there are the iBuyers, companies or platforms that use data and AI to make predictions of specific real estate markets and then make attractive offers directly to homeowners on off-market properties. 

The big names here are already quite well known: Opendoor, RedfinNow, and Offerpad.

Industry giant Zillow had it's own program, "Zillow Offers", which shut down in 2021.

AI For Real Estate Agents & Brokerages

Next we’re going to take a look at how AI has been impacting real estate from the perspective of real estate agents, teams, and brokerages.

This sphere of real estate has long been a hybrid of cutting edge technologies and good old fashioned communication. On the one hand, a large portion of homebuyers today start by searching for homes online, but on the other, people tend to use agents recommended by friends and family when selling their homes.

Agents constantly have to improve along multiple fronts, managing their online presence, working with increasingly sophisticated forms of media, all while honing their interpersonal skills and maintaining a variety of professional relationships. As such, any shortcuts or time-saving measures are usually met with some enthusiasm, and many have been jumping aboard the AI train for some time.

So let’s look at some uses of AI in real estate, both obvious and not-so-obvious…

Google Ads and Organic Search

This may seem odd at first. Afterall, when we run an ad campaign or try to rank on Google search, we aren’t the ones using AI - Google is!

And that’s true. Google Ads utilizes machine learning algorithms to analyze user data and optimize ads for better performance. And Google search uses AI to provide users with more relevant search results based on their search history and preferences.

As Google search and ads refine their targeting and understanding of users, real estate agents can take advantage of these processes, analyzing the data they generate to gain a better understanding of their target audience and adjust their marketing strategies accordingly.

For example, if Google Ads shows that a particular type of property is popular among a specific age group or demographic, real estate agents can tailor their marketing efforts to focus on that group. Additionally, analyzing search data can reveal important trends and patterns in the real estate market. This information can be used to identify emerging markets or target areas with high demand.

Overall, the AI and machine learning processes used by Google Ads and Google search provide data-driven insights that enable real estate companies to better understand their customers, reach their target audience and market their properties more effectively.

If you want to leverage this power in your own real estate marketing efforts, get in touch with us here at REW for industry-leading PPC advertising and SEO services. 

Chatbots

The widespread use of support chatbots on real estate websites was a relatively early form of AI-powered customer service technology that provided basic information and assistance to users visiting the site.

While they were often limited in their ability to understand natural language and provide nuanced responses, website chatbots were still helpful in providing basic assistance and directing visitors to the right resources.

These chatbots allowed real estate companies to offer round-the-clock customer support and answer common questions in real-time, such as providing property details, scheduling a showing, or connecting with an agent. They were also a new avenue for generating leads.

(Note: These chatbots are very different from current chatbots like ChatGPT which are trained on vast amounts of data using much more sophisticated methods. See below) 

AI & Real Estate: The Big Players

How have real estate industry giants been putting AI to use in their services?

Being some of the few with access to sufficient expertise, resources, and data, large real estate listing platforms were some of the first to leverage the power of artificial intelligence to enhance their services and stay ahead of the competition.

Here is a rundown of some of the top listing platforms in the business, along with some info about their AI endeavors:

• Compass - Along with standard search-based algorithms, Compass has also developed AI for creating detailed profiles of homes that are likely to sell in a particular area by examining potential home prices and comparing them with the local market conditions.

• Redfin - Redfin has two main tools powered by machine-learning. Redfin Estimate calculates the market value of any given home, taking into account over 500 data points. Redfin Recommendations is a newer product that provides customers with a list of recommended homes as well as notifications when a home they might like appears on the market.

• Trulia - Trulia streamlines home searches by suggesting properties based on a user’s unique preferences and search criteria as well as their similarity to other users. They have also integrated computer vision systems which can assess a home’s features based on photos, using these to enrich its matching criteria to show users the homes they want to see.

• Zillow - One of the most well-known AI-powered tools in real estate is Zillow’s “Zestimate” feature. This tool estimates a home's market value by analyzing various factors like location, size, and recent sales. The tool is regularly updated and serves as a starting point for buyers and sellers to understand a property's value. Zillow's neural network-based model allows it to update its estimates frequently with a relatively low error rate. Zillow also uses an AI-enhanced CRM that helps to qualify leads and find properties that they’d be more interested in.

The adoption of AI technology by these large real estate platforms has clearly worked for them as they now tend to dominate their respective markets. However, the landscape is beginning to change as smaller players are gaining access to technologies that enhance their services to better compete with their larger counterparts.

What has changed in AI?

While AI has typically been associated with human-like tasks, what makes AI so useful and powerful is its potential to do those tasks better than humans in one or more ways. And current developments in AI are no different because they also involve doing things at bigger scales, incorporating more data and producing results faster.

So what's changed? Why all the buzz around things like Google's Bard or Microsoft's New Bing?

Here's the difference...

Previously, AI tended to be reactive, domain-specific, dependent on very specific inputs, and limited in memory and learning capacity.

In contrast, new AI technologies are increasingly effective at complex tasks like image recognition and natural language processing. These new capabilities are due to a few key breakthroughs in AI technology, some of the most important being:

• Machine Learning - This is a subset of AI that allows machines to learn from data and improve their performance without being explicitly programmed. Machine learning is an iterative process where machines improve their performance in a specific task over time with little or no programming or human intervention.
• Deep Learning - This is a specific type of machine learning used to tackle complex patterns in data. It is particularly effective for tasks such as image recognition, natural language processing (NLP), and speech recognition.
• Natural Language Processing - NLP is the ability of computers to understand and process human language as it is spoken or written. NLP involves developing algorithms and models that enable machines to analyze, interpret, and generate natural language data, including text, speech, and other forms of communication. NLP technology has progressed in leaps and bounds in recent years thanks in part to new advances in transformer methods and the increasingly large sets of training data.
• Language Transformers - A transformer is a type of computer program that can help machines understand and generate human language. It works by looking at words and phrases in a sentence and figuring out how they relate to each other to form meaning. According to Wikipedia: "Transformers were introduced in 2017 by a team at Google Brain and are increasingly the model of choice for NLP problems."

The efficiency and speed of these methods has also increased the amount of data that can be incorporated into training AI. 

"When you go from 10,000 training examples to 10 billion training examples, it all starts to work. Data trumps everything." - Gary Kasparov quoting a Google engineer

These developments have unlocked some new and exciting features for AI, including:

• Conversational interactions- Being able to understand a wide variety of prompts delivered as a human would in the course of a natural conversation.
• Flexibility - Where older AI were more suited to perform one or a few tasks within a narrow domain, newer AI are capable of performing a variety of tasks (many of which are completely novel) across multiple domains.
• Learning ability - Newer AI chatbots are very good at taking in context clues and even learning from individual users over the course of their interaction. This makes them even more flexible and responsive. 

However, some questions remain: Will more sophisticated tools forever be inaccessible to individuals and small businesses due to a lack of access to data, storage, or processing capability.

And perhaps most importantly: Will AI just allow us to do the same old things better, OR will it bring about a sea shift in how the real estate business works?

 

Join the conversation and share your thoughts with our community in the Real Estate Webmasters forums.