Identifying the Right Opportunities: The Backstory
Taking a look at the past decade, we noticed a trend that 5M+ properties sell each year, which accounts for roughly 5% of the total commercial real estate market across the U.S.
Real estate transactions are complex processes, largely determined by geographical factors, economic conditions and personal decisions.
“How can we best target properties with a high likelihood to sell?”
That is a question we have consistently received since the company’s founding in 2013. We find that many of our users are interested in getting ahead of their competition and that identifying properties likely to sell in the future positions them ahead of the curve.
For a property on the market, the time between listing and closing can span several weeks or several months.
Even before a property listing occurs, the seller can spend months to finally decide to sell and prepare for listing.
In an ideal state, a real estate broker would be able to get ahead of this by efficiently identifying a potential seller before a selling decision is publicly announced.
Conventionally, this often involves the broker reviewing an entire list of properties and applying some heuristic rules to manually screen them. Aside from the inherent laboriousness, such a manual procedure heavily relies on intuition and experience, and is inevitably prone to biased interpretations and decisions.
Over the last few years, Reonomy has responded to this question with a simple rule of thumb: filtering properties with no sales in the past 10 years will help you identify properties with a higher likelihood of being sold.
While we believe this simple filter search can help unlock many adequate opportunities and assist our users in striking a deal, we were convinced that we could better support them in this task. We could leverage the property intelligence on 52M+ properties across multiple years to better identify these properties.
Today, Reonomy is excited to announce the launch of our latest Likely to Sell indicator to improve our users’ experience when looking for transaction opportunities.
Likely to Sell: What does it mean?
Market opportunities vary depending on the geographic area you are looking in. Market history has been different, and industries and demographics change over time.
Defining a general rule for identifying properties that will sell in the near future might be well suited for some cases but can exclude hidden opportunities in others.
This situation is best described as a complex interaction between neighborhood-level trends, history of a property and also a property’s intrinsic attributes.
For example, sales history needs to be taken into account while also considering the asset type, combined with the current situation of the market the property is in. All of these factors will be compared to the tax and debt history on a specific property to ultimately find ownership information to determine if the property will sell.
Some of these factors may be easy to pull together in a small geographic area, but it can be quite difficult to solve this intricacy for all CRE properties across the U.S.
This is where Reonomy Machine Learning comes into play. Our new methodology captures trends in the market on a national scale and works with local county level geographic attributes of the market in order to assign a likelihood score for each property to be sold within the next 2 years.
While it is difficult to pinpoint all of the reasons that drive the selling decision of a property, by leveraging more than 30 years of sales data on 52M+ properties across the U.S, our Machine Learning model is able to capture the owners’ predicted response to the subtle geographical and historical market changes surrounding the property.
This way, our unique technology identifies historical trends inherent to specific Zip Codes and Metropolitan Statistical Areas (MSAs) spotting areas in the country with higher or lower selling rate compared to the overall market.
This provides our users with a unique likelihood score for each property assisting them in ranking and prioritizing opportunities on the market.
A User Case: Identifying properties likely to sell in New York City
To explain how this new feature will be used in our platform, we will walk you through the journey of Joel, a Sales Broker interested in finding hot properties on the market in New York City.
Starting today, Joel will be able to use our platform to directly prioritize the properties he is interested in contacting for a potential deal.
Within the sales tab, Joel will encounter a new field, indicating a property’s likelihood to sell. Using this filter, Joel will have the opportunity to narrow his search results to potential deals with a high probability of selling in the next two years.
Now, the new properties cards will have a confidence level score corresponding to Joel’s criteria. He can then sort his properties by Likely To Sell score and start contacting the owners.
Joel is able to prioritize his outreach to these owners and build his off-market pipeline leveraging the Likely to Sell score.
Our new Likely to Sell feature will help to streamline your research process. Along with location and asset type filters, you can leverage this property intelligence to save time and make your discovery phase more efficient.
We look forward to offering additional machine learning functionality, empowering you to make more informed decisions and have an advantage over the competition.