Artificial Intelligence in Real Estate

18 applications and main challenges

While many industries have benefited from Artificial Intelligence, Real Estate is essentially a virgin territory to apply AI. This blog post will show you how you can transform Real Estate with AI, the key consumers of these solutions, and the challenges the industry faces in its adoption.

Suppose you are running a Multiple Listing Service, a Real Estate agency chain, an investment fund, or you simply want to create a startup in the industry. In that case, you will discover dozens of use cases to optimize your Real Estate operations with Artificial Intelligence.

Course, Templates

Data Ignite

Build your company data roadmap using our Data Ignite methodology

Learn More

Who are the key consumers of AI in Real Estate?

The primary consumers of Artificial Intelligence models in Real Estate lay in four groups:

  • Construction companies: looking to shorten the construction time and costs.
  • Investors (professional investors or retail ones): understanding the potential of an investment.
  • Realtors: managing leads in a personalized and optimal way.
  • Property managers: improving the management of tenants’ contracts and streamlining property maintenance.

What are the main applications of AI in Real Estate?

Each of these groups optimizes for different targets at different stages of the property lifecycle. As such, they require various forms of support AI can provide. Therefore, we will divide the applications according to their target user:

AI application for Construction

  • We’ve seen a growing demand for Computer Vision technologies that support the digitalization of construction blueprints. Why? Mainly to automate error-prone and resource-intensive tasks such as budgeting (estimating materials, timelines, etc.).
  • Another area where Computer Vision can excel in Real Estate can be found in property inspection, compliance, and work safety. AI models can help you identify risks that you could neglect otherwise.
  • The ultimate goal is to produce personalized designs that optimally fit the end-user needs. These designs should consider explicit requirements (e.g., number of rooms, bathrooms, etc.) and overall efficiency (i.e., construction cost, energy efficiency, etc.). Some research has flourished in this area. However, we still consider it a long run until we see genuinely autonomously designed buildings. In the meantime, let’s use AI to support the design, providing key metrics that could help them make better decisions and iterate faster.

AI applications for Investors

Here, we will include both buyers and sellers. As well as professional investors and homeowners. Most of the AI applications for Real Estate investors consist of accurately estimating KPIs about individual properties and markets. Examples of such indicators include:

  • Automated valuation of properties: so the investor knows what would be a fair value for that property.
  • What would be the After-Repair Value after doing X? So, the investor can make an informed decision about the cost-benefit of each renovation.
  • Key financial indicators in their various colors and flavors: ROI, gross yield, net yield, appreciation, occupancy rates, rent value, etc.
  • Forecasting the dynamics of a particular market. For example, identifying which markets are gaining demand, where’s the supply falling short, etc.
  • Supporting the buying/selling process: realizing the trade-off between the asking price, offer, and likelihood of closing the deal.

AI applications for Realtors

Realtors are salespeople with a steroid charge of street knowledge about Real Estate. As such, the applications of AI for realtors are similar to those observed in general marketing and CRM for optimizing the customer journey. You can learn more about it in our Youtube video: 11 AI use cases in Marketing to improve your customer journey. In summary, AI can help realtors optimize the overall sales funnel:

  • Lead scoring.
  • Personalized recommendations.
  • Identifying churners (sellers that will leave for another realtor).
  • Identifying buyer capacity (and willingness to buy).
  • Reaching each lead at the right time, through the right channel.

In 2021, we won an open innovation challenge promoted by Vonovia and the Hands-on Data Initiative. Vonovia is one of the largest European companies, with over 500k units. In the challenge, we worked on a recommender system (called Flat Finder) to help Vonovia suggest better-suited flats to their leads, improving the overall customer experience and minimizing the time to find the right apartment.

AI applications for property managers

On the property management side, most use cases rely on forecasting events concerning the property (e.g., damages) or the contract with the client (e.g., churn). As such, you may think of AI applications for property management, such as

  • Predicting churn at a tenant level. Ideally, such a prediction should respond to a specific commercial action the owner wants to apply. For example, what’s the churn probability if we increase the rent? Or in case we only renew the contract for one year?
  • Predicting which customers will most likely have delays on their monthly payments is a relevant application where AI can support property managers and investors.
  • Predicting occupancy rate versus pricing (especially relevant for short-term rentals).
  • The likelihood of certain damage arising from a unit. Alternatively, how much is a property expected to demand money-wise soon?
  • Also, there’s a lot of potentials to optimize the workforce for tenant complaints, from identifying the root cause to identifying the best team to solve the problem and the best time to visit the property.

In 2023, Vonovia launched three open innovation challenges through Hands-on Data. All of them with a focus on optimizing the management of properties. We won the Image Insight challenge, triggering a new collaboration between NILG.AI and Vonovia. In the challenge, we want to identify and characterize the damages on a tenant request, understanding who is the right team to proceed with the work and what replacement parts would be required. The other two challenges focused on optimizing the workforce for renovation (i.e., understanding delays, reducing bottlenecks, etc.) and optimizing the energy efficiency of the apartments through dynamic control of their heating systems.

Do you want to further discuss this idea?

Book a meeting with Kelwin Fernandes

Meet Kelwin Learn More

Why is Real Estate falling behind in AI versus other industries?

We have worked with several clients in the industry. However, as a whole, we haven’t seen a major adoption of AI in Real Estate. There are four main reasons Real Estate hasn’t fully benefited from Artificial Intelligence compared with other industries:

Lack of reliable open data.

Most open data comes from Multiple Listing Services (MLS) such as idealista and imovirtual in Portugal, which isn’t entirely reliable. We face problems such as duplicate listings, low-quality entries, and, most importantly, the data is about asking prices but doesn’t consider the final transaction value. So, data is biased and represents an overestimate of an essential value in Real Estate: price.

Market Fragmentation

In most markets, the industry is highly fragmented, with most players being individual homeowners and small investors, and just a few transact and hold a sufficient amount of properties. With limited actions, the amount of data you can collect and the profit you get from internal optimizations will be limited. So, more creative ways of monetizing AI must surge in such contexts. We talk about it in our AI monetization video

Multiple listing services have a dominant position here, since they can benefit from the scale they get through the aggregation of massive amount of data and access to end customers. So, the suit spot to run AI models in Real Estate is at MLS.

Each and every property is unique.

Each property has its own unique properties that separate it from the rest. It’s difficult to quantify what makes two properties comparable at scale properly. You can consider putting into your model’s variables such as size, construction year, number of bedrooms, etc. However, getting street knowledge from specific markets isn’t trivial. The distance from a large employer can be a critical factor in a particular neighborhood. But, in the next neighborhood, the main driver of property value is avoiding a street with massive traffic jams. At least, it isn’t trivial at scale.

So, street knowledge is crucial if we want to compete with local stakeholders, and we need ways of creating relevant human-machine interactions when assessing properties.

Experiments can be expensive

Training AI models tend to require time and iterations. AI models produce dumb results and get better as they gain experience. Each learning iteration in Real Estate is extremely expensive. Letting an AI discover new ways of building a house just to realize it’s suboptimal to re-train the AI isn’t very cost-effective. So, we need to either learn from historical data or let AI be more of an enabler of the human process than an autonomous decision-maker.

How to benefit from the current Real Estate landscape with AI?

Given the landscape we just depicted, there are two possible scenarios for using Artificial Intelligence in Real Estate.

  1. You can benefit from AI if you are a large company operating in the industry. So, you can benefit from the large-scale optimizations on any stage: building, investing or managing your Real Estate portfolio.
  2. Alternatively, you should consider becoming a service provider for the industry, where you can serve the entire industry with an AI-based product.

In any of these scenarios, we can help you find and execute AI solutions that succeed in this industry.

Like this story?

Subscribe to Our Newsletter

Special offers, latest news and quality content in your inbox once per month.

Signup single post

This field is for validation purposes and should be left unchanged.

Recommended Articles

Business-centric AI: A New Perspective for Your Company

Coping with the challenge of integrating AI into your business? You’re not alone. Many companies struggle to find the right approach to AI, often getting lost in technical details or data management issues. However, there’s a solution that transcends these common pitfalls: Business-centric AI. This transformative strategy is the perfect way to align your core […]

Read More
Long-term vs. Short-term Predictions in Machine Learning

When building a machine learning model, one of the most common questions is whether to opt for long-term or short-term predictions. In other words, should you build a model that forecasts an event tomorrow or a month from now? Our article will demystify this critical decision-making process. We’ll walk you through a strategic approach that […]

Read More
Ditch the Crystal Ball: Reverse-Engineering with Machine Learning

  Machine Learning models are estimators – which means they can be used not only to predict unknowns in your business but also to reverse-engineer complex business processes. As part of this blog post, you will learn how to identify these potential points of improvement, prioritize them, and create models to estimate them. Identification How […]

Read More