There is a lot of paperwork involved in conducting business in commercial real estate.
And while that may sound like an annoyance, for brokers, investors, lenders, appraisers, and service providers, it actually points to a fountain of public information that serves up insights on properties and owners.
The thing is, due to the number of sources nationwide, CRE data has always been wildly unorganized and strewn about. The data has always existed, but has seldom been brought together in an accessible, cohesive manner.
Enter: entity resolution, changing the data game completely.
In this article, we’re going to take a deep dive into entity resolution. Through defining it, explaining how it works on a high level, and using Reonomy’s own data solutions as an example, we’ll describe how entity resolution is enabling millions of data records to be more cohesively structured, more accessible, and thus more powerful for everyone in CRE.
Think about this…
For every property transaction, mortgage, lease, valuation, or building service that takes place, there is a contract and/or report that needs to be filled out and filed.
In those reports are details of the transaction or event that took place, from initial agreement to final payment—including the parcel, people, and amount of money involved.
So, from a data standpoint, there is plenty to be taken advantage of.
The things is, however, those reports are recorded by real people—not to mention by different people in each municipality. That brings about two distinct issues from a data point-of-view:
- Inaccuracy: Human error is natural, and so certain points of information will be recorded inaccurately.
- Inconsistency: Given the number of different people, processes, and offices involved across the country, records will be inputted and written in different ways, and therefore will not always match from one source to the next.
Furthermore, since CRE jurisdiction and recording sits on a state or county-level, there is lacking formality to the way records are written, exacerbating the issue of inconsistency.
Take a property address, for example. It is one of the simplest pieces of information on a commercial property, yet, given the amount of people involved and the lacking formality from county to county, an address might be recorded in a variety of ways across different data sources.
Consider the many ways a single address could be written:
- 29 9th Ave, New York, NY 10014
- 29 Ninth Ave, New York, NY 10014
- Twenty Nine 9th Avenue, New York, NY 10014
- 29 Ninth Avenue, New York City, NY 10014
The list goes on…
While we read these all the same, that does not mean that the technology collecting data do, too.
And, mind you, that is only for a single data point on one property.
Project this same idea over hundreds of data points on each and every property and owner in the nation’s 3,100 counties, and the messiness of commercial real estate data can get a little out of hand.
Bring entity resolution into the equation, however, and the narrative of commercial real estate data takes a much more positive turn.
What is Entity Resolution?
Entity resolution is a computer intelligence process that takes disparate data points, refines them, and links them together under a single unique identifier.
It is an algorithmic “task” that matches, deduplicates, and fuses separate data points together that all have the same subject or “entity.”
What is an entity in a database?
An “entity” is simply a synonym for a single subject—the pillar, the nucleus, the commonality among a variety of data points.
In the case of CRE, an entity might be an individual person, property, or company—it is essentially a noun that you want to collect or store multiple pieces of data on.
A database is obviously going to have many subjects or entities, with multiple data points on each entity.
For the sake of example, however, consider just a single property. Let’s say one entity in your database is the property we mentioned above—29 9th Ave, New York, NY 10014.
Perhaps your database consists of property addresses and building and lot measurements, and you use separate standalone tools to find owner contact information and historical property sales data.
With entity resolution, you can take the data from all of these sources and bring them together into a single cohesive dataset.
In this case, your existing and incoming data points on 29 9th Ave, New York, NY 10014 can be seamlessly fused together and deduplicated, no matter what source they’re taken from. That includes taking the address from each source, even if written slightly differently, and linking it to the same, singular data entity.
That also includes taking sales data, ownership data, mortgage data, and anything else, and attaching that to the same, singular data entity.
One entity, with a variety of data points.
You’re left with more depth and less noise attached to every entity in your database.
Instead of new data being fed in to an existing database and essentially poured on top of what already exists, it ‘s worked into and matched to what is already there.
A Real-Life Analogy of Entity Resolution
Let’s take a big step back and look at a very, very simple analogy. While the example below does not represent the sheer magnitude of entity resolution as applied to data, it can be used as a simpler way to understand how the process works on a very high level.
So, consider this… There are 13 numbers and royal figures in total in any deck of cards (A, 2, 3, 4, 5, 6, 7, 8, 9, 10, J, Q, K), with four suits for each, equally 52 cards in total.
Consider each number or royal figure to be one data entity, and each suit to be one data type (like, say, property ownership, sales data, or mortgage data). An entity could be a single property or property owner.
One physical deck of cards represents one data source. So, to recap:
- One deck of cards = One data source
- Each number represented = One data entity in your database
- One suit = One data category/type
Again, for the sake of example, let’s consider a single entity—the number 7.
In any one deck, you have four 7-cards. A 7 of hearts, clubs, diamonds, and spades. If you have five decks (or five data sources), you’d have twenty 7-cards in total, some duplicates, some not. Make sense?
So, let’s say you’re looking for a 7 of Hearts, specifically. Across five decks, you would have five 7-of-Hearts cards, right?
Each source, or deck, would be giving you a different version of the same thing—a 7-of-Hearts.
Instead of taking all of those 7-of-Hearts cards and simply dumping them together into your “database,” entity resolution is equivalent, in this example, to someone sifting through each deck of cards, seeing that there are five 7-of-Hearts’, stapling them together, and labeling them under a single identifier.
If you’d like to add more breadth to your knowledge on the 7-card, you could do the same for the seven of clubs, spades, and/or diamonds. In the end, you’d be left with only four cards (or four data points) on the 7-card, instead of twenty in total, with duplicates of each.
Now, back to CRE. This is where Reonomy comes into the equation.
Reonomy Entity Resolution
Whereas the above analogy represents the idea of entity resolution across just 52 entities, Reonomy’s entity resolution and data solutions do as much (and more) for over 50 million multifamily, land, and commercial properties nationwide.
Not to mention that there are much more than just four data points on each of those 50+ million entities.
Reonomy’s database includes sales, mortgage, ownership, tenant, and tax data on individual properties, linked together and cleansed with the help of proprietary algorithms.
These algorithms enable Reonomy’s entity resolution capabilities. They are what enable Reonomy to have the biggest commercial real estate database in the country, and are what Reonomy’s search app is built on.
Reonomy’s entity resolution capabilities also allow for our data to be merged and integrated within any other database, of any size.
While Reonomy’s database has its own value due to sheer depth and breadth, the aforementioned proprietary algorithms have standalone value in being the engine that companies can attach to their own database to clean, dedupe, and fill in their existing data, while also adding in whatever new data they seek from Reonomy’s database.
So, in the real-world case of Reonomy, the number 7 could be looked at as a single commercial property. Hearts, spades, clubs, and diamonds could be looked at as sales, mortgage, tenant, and ownership data.
And whereas our example laid out the process for a single entity, Reonomy gathers, hypothetically staples, and labels layers of data on more than 50 million commercial assets.
Visit our Data Solutions page to learn more and/or inquire.
What does Entity Resolution mean for commercial real estate?
On a very high level, the benefits of entity resolution are pretty straightforward—cleaner, more organized data and databases.
Entity resolution allows for new levels of automated data maintenance, including:
- Data Cleansing
- Record Linkage
- Database Scaling
Entity resolution allows for data to be much cleaner for two reasons. First off, because duplicates are removed.
Secondly, since matching data points are being linked together, the likely accuracy of each data point can be better assured based on its consistency across various data sources.
This one is pretty straightforward. No more loose data. No more duplicates. Record linkage essentially takes all of the noise out of a database by consolidating and fusing all of your incoming data from any number of sources.
With the ability to merge data records, comes the ability to scale the size of a database.
For example, Reonomy basically works as a filter that you can attach to your database that takes all of your incoming data, deciphers where it should go, and fuses it with existing records as necessary.
Instead of just piling new data onto existing data, entity resolution allows for your data to be efficiently and cleanly scaled in terms of both depth and breadth. For example, you could either add more data entities to your database, or you could add more data on those entities. Your coverage can grow, and your depth of knowledge can grow.
Entity resolution takes out the concerns of adding an additional data source, as well. With no incoming noise, it is much easier to continuously add new data to your database.
It’s what comes from this higher quality data and scalability, however, that shows the true benefit of entity resolution as related to commercial real estate companies and individuals.
Benefits for CRE Companies & Organizations
When it comes to brokerages and other large commercial real estate firms, entity resolution allows for much more efficient processes on a company-wide scale.
The time it normally takes to compile the information needed to research and analyze properties and owners, especially in bulk, can be a major hindrance to efficiency.
With data solutions like Reonomy’s, every individual across an entire organization can begin saving time immediately.
Entity resolution can help build a database that creates better organization-wide workflows.
Decisions can derived much quicker, not to mention the fact that those decisions will be better-informed. It is something that can be harnessed to arm your team or organization, better enabling them to research properties and owners, glean powerful insights, transact properties, service properties, and so on.
Benefits for Individual Professionals
On an individual basis, or say, if Reonomy data is accessed through the Reonomy web app, entity resolution is still largely in-play.
Entity resolution allows for individuals to access more refined data that can not only help them make better, more-informed decisions, but can also save them a ton of wasted time that’d normally go towards fumbling around with inaccurate data.
In the end, entity resolution is a functional data refining process that works at the core of the way entire organizations function. Reonomy’s entity resolution capabilities along with the biggest CRE database in the country allow for companies and individuals to operate at higher planes of efficiency.