HelloData: AI Powered Multifamily Rent Comps

Published by
Marc Rutzen
on
February 26, 2024
HelloData: AI Powered Multifamily Rent Comps

Automated Rent Comp Detection

Rent comp identification can be a time consuming and uncertain process. Real estate investors, brokers and appraiser do the best they can with limited data, but subjectivity and lack of access to real-time data can easily lead to subpar results.

HelloData takes rent comp identification to a whole new level. We use AI to analyze EVERY property in your market and automatically recommend the most statistically similar rent comps:

HelloData UI: Automated Rent Comp Detection

Our similarity score (shown above) takes into account several attributes to ensure the best comps are selected every time:

Submarket

This variable accounts for the demographics and supply & demand data in the hyper local market surrounding each property (down to the census tract level). We account for median income, median home value, school district ratings, crime, average rents and time on market, unit mixes among similar properties, and several other variables to get an apples-to-apples comparison of the market immediately surrounding each property.

We did this to account for suburban markets where you may not even have another Class A multifamily comp in the same market if there hasn't been new development recently, and to handle markets like Chicago where neighborhoods can change significantly from one block to the next. Our algorithm can jump to the next comparable market to identify properties that truly compete for the same residents, which is a major advantage to our approach.

Distance

We account for physical distance, but it's actually a proxy for density. If the market is very dense, there will be more comps in a tighter radius around the subject property. Going a few more blocks really can and should decrease the probability of a property being considered a comp. In less dense suburban or exurban markets, the algorithm allows for miles of separation between comps.

A simple radial analysis really doesn't work well across the board because markets differ in density. Our system automatically accounts for these differences across markets.

Number of Units, Vintage and Stories

Any approach to identifying comps should take these variables into account, and ours is no exception. Instead of using hard filters to remove comps though, our system weights these variables alongside every other attribute to get a holistic sense of comparability. When used alongside the other variables we analyze, this approach produces empirically better results.

Quality and Amenities

This is where we really differentiate in our approach to comps. Our platform uses computer vision to analyze photos across millions of listings every day, detecting the room type, condition and quality to assess comparability from the resident's perspective. This approach helps us pick comps with a similar look and feel... without having to manually look at millions of listing photos.

We also aggregate and standardize data on over 200 different building and unit level amenities to gauge the statistical overlap of amenity sets, producing much more accurate comp selections:

HelloData UI: QualityScore Listing Photo Analysis

With each of these innovations, we produce rent comp recommendations that overlap with appraiser selections 9/10 times. We've proven this by comparing our rent comp selections to those in actual appraisals provided to us by our clients.

Want to learn more? Schedule a demo and see how HelloData can reduce your rent survey times by over 80%!

Marc Rutzen

Marc worked in real estate for 5 years before launching multifamily analytics startup Enodo, which he sold to Walker & Dunlop (NYSE: WD) in 2019. At W&D, he served as Chief Product Officer, developing products that helped source billions in loan volume. Outside of work, he enjoys reading, running, and spending time with family.

Recommended Articles