Here at MapVida, we spend a good amount of time touting our similarity models — where people can find “look-a-like” neighborhoods and the ability to “build” an ideal area. We go back 17 years, looking across tens of thousands of data points per neighborhood, and determine – both overall similarity and across 7 dimensions (more on that in a moment).
Why did we create overall similarity and different dimensions of similarity? First, it sounds cool. But if we are going to allow people to build an ideal neighborhood, and create a decision-making tool that people can actual use, we had to create a layered model…one that provided for the ability of someone to say, for example “I care more about walkability than I do about school quality.” That, and by grouping key points of interest about a neighborhood together, people can more easily compare neighborhoods to either another location, or to a city average.
First, whenever we compare two areas, we show how similar they are across the seven different dimensions (Housing, Lifestyle, Schools, Safety, Walkability, Outdoors, and Businesses). Sometimes we provide additional context — like how these areas also compare to a city index. Let’s walk through examples:
This is a big dimension of information. We look at who is living in the area, what do they do, what do they spend their money on etc…etc… and then we create comparisons of these different datapoints to help explain why we think two areas are similar or not.
With this dimension, we look at owner/renters, housing sizes, types of housing (single family, duplex, triplex, multifamily), and affordability (rents and home purchases vs. city index. We then compare those data points between the different neighborhoods we’ve determined to be similar.
The walkability data equates to average commute times, plus access to entertainment, grocery stores, restaurants, and retail locations. We show context by comparing to the city averages and the other neighborhoods.
The Other Dimensions: Safety, Schools, Outdoors, and Businesses
Similar to the other dimensions, we look at crime (personal, property and overall), school quality (pre-school through high school), outdoor access (parks, water access), and local businesses in the neighborhood. Again, we index to the city averages where it makes sense to provide context.
It’s human nature to want to compare things to what is already familiar, and we know that people use comparisons to make decisions all the time. But that fact is that humans aren’t very good at being accurate or unbiased, and that’s where the data comes in. After 2 years and 3 iterations of our models, we have created tools that make sense of millions of data points to allow anyone to compare neighborhoods based on the factors that are most important to them. And businesses can use this same data to understand where their best customers are and how to replicate past successes.