What is the “similarity” score?

Our rent comp detection algorithm accounts for distance from the subject, building amenities, unit amenities, number of units, stories, vintage, and finish quality (derived from QualityScore) to surface properties that are statistically very similar to the subject.

Since our approach leverages AI, the weights of each variable differ by market – accounting for urban density or suburban sprawl by automatically changing the weight of variables like distance and number of stories. Because of this, we don’t have fixed weights for each variable, but we do show an objective measure of the percent similarity for each attribute. These individual attributes combine to create an overall similarity score for each of the rent comps relative to the subject.