Data is the lifeblood of property analysis. From local demographics to employment statistics, the real estate industry has long relied on established metrics of conventional analysis to evaluate multifamily assets. But are traditional methods of analysis truly delivering the greatest possible value to the industry?
In a world being transformed by advances in data science and artificial intelligence, there’s a veritable treasure trove of untapped data sources that can supplement our analyses and ability to predict asset behavior — and thereby unlock new value for the real estate industry. Among the hundreds of data sources that can deliver deep, value-enhancing insights are some unexpected and even exotic ones. With billions on the line, here’s a look at what our firm has uncovered about the unconventional data sources that the industry can’t afford to ignore in the era of AI.
Service Calls
Want a better idea of what a neighborhood is really like? You could do far worse than to examine publicly available service call data. Service complaints about issues such as utilities problems or road debris can provide a telling view not only into a neighborhood’s standards of infrastructure maintenance as a whole, but also the most common issues facing a neighborhood. This, in turn, can tell us a great deal about property values in the neighborhood. Significantly more complaints concerning noise levels than broken sidewalks, for example, is indicative of a higher-level neighborhood. The service call metric can be used to analyze trends both negative and positive, helping predict up- or downswings of a region.
Health Food Stores
Trendy health food supermarkets signal more than just a local appetite for fresh, quality foods at premium prices. A property’s distance from the nearest high-end grocer also tells us a great deal about its likely value. Crossing real estate asset value data sets in a given region with the distance of each asset from the nearest Whole Foods reveals to us that as the distance to the nearest store gets longer, asset prices get lower. Whole Foods clearly picks its locations based on positive criteria such as high income and education levels.
Street Trees
That tree-lined streets are more appealing to residents should come as no surprise. But do they really tell us much about the desirability and overall value of an area? Data science answers with a resounding yes. In dense urban centers, green spaces are often few and far between. A tree-lined street is evocative of an oasis that gives residents some much-needed nature within a gritty environment and suggests a well-maintained neighborhood.
You can extract the tree levels for a given area either from satellite images, or, if you don’t have access to satellite analysis software, from the street tree numbers and locations data for major cities like New York and San Francisco, which have made this information publicly available.
Asset Names
Property names affect value significantly more than you may have thought. It turns out that particularly when it comes to multifamily, certain names are “better” than others, and a property’s name can carry a rather strong correlation with its price level compared to similar assets with less-valuable titles and associated descriptors. Assets with names like New York, Hamilton, Villas, Residences, by Windsor and Lofts tend to be of higher value. Meanwhile, properties associated with words like Astoria, Raintree, Pangea and avenue are on the lower-performing end of the spectrum.
Flight And Taxi Statistics
Is a city seeing a sustained increase in incoming flights? That can be a sign that the area is heating up — and is ripe for real estate investment.
Simple flight databases can be found online. Flight stats can then be correlated with other data sets, such as visitor demographics and current real estate trends in the areas they’re coming from. Local taxi activity is also an illustrative indicator of an area’s vibrancy. Seeing an uptick in taxi drop-offs at leisure spots or nightclubs? You’re likely looking at an area on the upswing. By monitoring such data, you’ll be well-positioned to obtain early insight and beat the market to prime opportunities.
National Oceanic And Atmospheric Administration (NOAA) Weather Data
It’s not exactly news that certain weather factors can influence real estate valuation, but the relationship isn’t as clear-cut as you might think. For example, if you want an accurate characterization of a region’s property values, you’ll need to do more than merely count NOAA’s total of rainy days per year. Our firm engineered a new analytical feature that dives deeper, looking at the number of consecutive rainy days per year, and found that if the number surpasses 20, there’s a negative impact on real estate values.
Similarly, consecutive sunny days can also be highly illuminating. You might assume that places with higher consecutive sunny days carry higher property value, and that’s true, but only to a certain point: Overall sunny days are very common in desert regions, which tend to fall into a lower-cost category.
As AI and data science enable new efficiencies and value creation across a wide spectrum of industries, real estate stands poised to reap big gains from greater use of these tools. New data sources are rolling out every day, and as information becomes increasingly accessible through the internet of things (IoT) and other developing hubs of connectivity, we’ll see more and more unconventional sources put to use in future analytic evaluations — and exponentially more value unlocked for investors.
The new data-sourced methods of carrying out multifamily analysis are based on machine learning and access to abundant data, but at their heart is the creativity of human real estate operators and analysts. Almost any geographical or behavioral metric can reveal valuable predictions about the areas they’re in and the people carrying them out, as long as they are analyzed the right way and with enough data.