The HomeLight Methodology

How we find the best real estate agents

Starting with data

We believe that when considering one of the biggest financial transactions of your life, you should have 100% transparency in order to make the optimal choice -- something more scientific than a referral from a friend or a sign on a park bench.

With our unique data set, we apply statistical data analysis and machine learning methods to predict who will bring the most experience, expertise and negotiating power to your home transaction. The result: a list of the best agents personalized to your needs.

Gathering your input

With your input on the following categories we're able to better understand your requirements and challenges. Each piece of data provided allows us to match you with agents precisely suited for your needs. We use these inputs in the following ways:

LocationUsing spatial IP and geo-distance analysis, we can target real estate agents that have hyperlocal expertise within your city, neighborhood, and even those that know your street.

Buy or SellThere can be vast difference between a good listing agent and a good buyer’s agent (and very few are experts at both). We can filter agents by their historical expertise, buying or selling.

Property Type Work required to stage a home in a city is completely different than that needed to sell an empty lot. Our transactional data allows us to identify agents’ expertise in your exact type of property.

PriceAn agent familiar with your price range is a must. Luxury real estate agents can create exclusive showcase & selling opportunities. While agents that specialize in first-time home buyers can secure the home of your dreams on a budget.

Agent TypeHiring an agent is both an important financial decision and an extremely personal one. For some, having an agent who can spend a lot of time fostering the relationship is crucial. Others just want a tough negotiator, who can sell their home fast and for the most money.

HomeLight Matching Engine

The HomeLight matching engine analyzes over 18 million transactions and 2 million agents. There are two primary components to our machine learning algorithm:

Relevance Analysis RankingsEvaluates dozens of variables to find agents who have successfully advised clients similar to you.

Real Performance RankingUtilizes Bayesian statistical models to predict which agent will get the best results for you.

The final result is an objective list of the best agents across all of the major brokerages in your neighborhood. You can choose to meet with as many of these agents as needed.

Your Personalized Agent Recommendations

Once you’ve received your matches, you may view an agent's expertise and experience in detail. We have compiled a list of agent performance data that cannot be found elsewhere online including:

RankAgent ranking across different variables and versus other local agents

Geo SpecificityWe analyze the geographic expertise of each of our agent matches and show you their transcations on a map.

ListingsRecent listings this agent has completed

ExpertiseEducational background, proven local experience, and agent bios

ReviewsQualitative customer reviews by individuals who have interacted with this agent

VolumeAgent transaction volume by year

Find an Agent