GIS data can mean different things to different companies. Trulia uses GIS data to help consumers understand local information and empower them to make smart choices whether they are looking to buy or rent a home. When Trulia launched in 2005, the goal was to open up real estate data for the masses in a vertical that had traditionally been closed and difficult to navigate. We initially did this in a simple, yet clever way, by geo-coding property listing addresses and displaying the listings on a map.
Today, this type of process is quite straight forward, but when we launched it back in 2005, APIs like Google Maps were not yet readily available, so the project was innovative for its time. We had to build systems and develop expertise in order to achieve our goal. While it might have seemed like a big undertaking then; these early investments were a fantastic way to build our data foundation. And we quickly realized the value of having sophisticated GIS capabilities and the hunger that consumers had for the information we could produce from this data.
To start, we built upon the listings data collected through GIS to analyze market trends by geography. By focusing on the visual output of these queries to display map data layers, we opened up a fresh way for consumers to quickly understand trends in their neighborhoods and cities. By combining this data with listing locations and prices on a map, we created a brand new way for consumers to spot deals and understand how listing prices compared to neighborhood trends.
"More public data is being opened up every day, and by combining interesting public and proprietary datasets together, we are able to achieve our mission"
The success of those early endeavors led us down a path to build up our GIS with data from many sources. Since our GIS is focused on the United States, our core data hierarchy contains states, counties, cities, zip codes, neighborhoods, tracts, blocks, and parcels. We also layer on shapes to describe points of interest, like parks, playgrounds, hiking trails, schools, school districts, and public transit lines.This was built using public data, such as US Census TIGER and the General Transit Feed Specification (GTFS), along with proprietary sources for items such as neighborhoods and parcels.
To build upon this spatial hierarchy, we have added additional layers of attribute data. This starts with our own proprietary data, such as rental pricing and the year each home was built, which allows us to create a number of different maps, including our Affordability Map or maps to animate the rate in which communities have been built up over time. We also added in demographic data to help consumers understand who lives in a neighborhood. And, from sources such as USGS and FEMA, we have also added in layers to describe the likelihood of natural disasters as a result of forest fires, earthquakes, and flooding–you likely aren’t always thinking of these when house hunting, but this information is critical when considering your next home and the insurance needed. Finally, because we know living on a quiet street is important to a lot of people, we built data visualization models to show traffic volume on streets, alleys, and ways.
Given that many hopeful home-buyers have a family or are looking to start one, schools are another important data point to Trulia. We display districts and attendance zones on our core maps, and also use our GIS to pull out specific listings that fall within these polygon areas. We further this by adding rating information for each school and district on our maps to help demonstrate the quality of individual schools, the spatial area of each attendance zone, and the available homes for sale or rent, all in a seamless manner.
More recently, we have invested in data to demonstrate crimes and commute times, along with integrating the wealth of the amenity data that Yelp provides.
For crimes, we are able to classify the type and severity of individual crimes that are reported to local authorities. Meta data for these data points includes latitude, longitude, date, and time. To summarize this in a useful way, we use street information that we’ve built into our GIS to adjust crimes to the nearest major street intersection. This allows for anonymizing exact locations, and makes a consistent way to summarize the data in our spatial hierarchy. Our heat map layers allow consumers to quickly identify higher crime areas in a city or neighborhood, and at the parcel level we are able to assign a specific crime rating for every property in the United States.
Knowing how long it takes to get to work or other points of interest is another important criteria in house hunting, and our GIS allowed us to combine street data with a proprietary routing engineand GTFS data to create commute time estimates. Using a unique process, we use the GIS to pull enough data to calculate times from given latitude, longitude point in 360 degrees, assessing every available route via both car and public transit. We further visualize this with a choropleth map to help consumers easily understand which properties are within certain time bands from important places, such as where they work or the grocery store.
Aside from these important product features that we grant consumers free access to, we use our GIS in other areas of the company. For example, our economics team uses it when compiling housing statistics, such as in our annual Rent versus Buy Report, where we break down geographic areas with current pricing trends to determine whether it’s more cost effective to purchase or rent a home. Trulia’s public relations team also uses our GIS data to tell insightful stories that help consumers better understand real estate and trends in the industry.
At just over 10 years old, our GIS continues to grow and be the center point of how Trulia operates. More public data is being opened up every day, and by combining interesting public and proprietary datasets together, we are able to achieve our mission at Trulia of opening up data, empowering consumers to find a home, and helping them learn about potential areas they want to live in. I can’t wait to see what we do in the future, with even more data.