Research & Insights

By Seth Teicher, August 21, 2014

Mapping Adventure: A Data Driven Exploration of America’s Parks

I get outdoors all the time and am constantly dreaming of new adventures. While the National Park websites are a decent place to get inspired, it’s hard to find reliable information on state parks. Certainly not on one, easy to browse site.

Fed up with the online resources available, and keen to share ideas with friends before Labor Day, while also planning a trip for myself this fall, I decided to create the guide I always wanted. It took me less than a week. Using the crowd, I analyzed nearly four thousand state and national park websites, identifying all the awesome places I can hike, bike and camp across the country. 

While the Bay Area has tremendous options, I’ve been dreaming about a trip to Oregon or Washington since I moved here. Its beene nearly two years and I still haven’t made it north of the Lost Coast.

As most do, I began perusing the web to look for inspiration. To my disappointment, none of the popular directory sites captured the breadth of possibility that I knew the Pacific Northwest had to offer. Furthermore, when I started down the road of actually looking at the official park websites, I found them frustrating to browse and tough to compare my options. 


Frazier Lake State Park, Washington // Patrick M via Flickr Creative Commons

I knew I couldn’t be the only one who felt this way. So, given CrowdFlower’s unique capabilities, I decided to create the tool I wanted and see all my options on a single map. 

While the idea started with California, Oregon and Washington, by the time I was scraping Wikipedia last Saturday , it occurred to me that it’d be way cooler if I applied this approach to the entire country. 3,852 parks later, I give you the National and State Parks Adventure Map 1.0.

Visualized on Mode Analytics, the map lets you sort by activity and drill down into park metadata, such as the official website, TripAdvisor and Yelp pages.

While I’d initially intended to capture all the activities and facilities I could dream of, such as where I could go spelunking or scuba diving, or whether the park had power sources or canoe rentals, I settled on nine categories that generally captured the variety of things that each park has to offer.


Snowshoeing Mt. Rainier, Washington // Nickay3111 via Flickr Creative Commons

To accomplish all this, I first scaped Wikipedia for all the National and State Parks. Once the source data was assembled, I set up a series of data enrichment jobs to collect the following:

  • Official park website (or semi-official site from a database if the official site didn’t exist or couldn’t be found)

  • TripAdvisor park website

  • Yelp park website

With these resources in hand, I could now prompt contributors to research all the available activities for each park. And research they did. About four days of crowd work later, I had neatly structured and labeled data that was almost ready to go.

The last thing I had to do in order to visualize the parks on a map was to have the crowd collect latitude and longitude. The job was fairly straightforward. Via the source date, I auto-populated the park name and state into a Google Maps iFrame within the job. Once contributors confirmed that they saw or could find the park, they dropped a pin and easily retrieved the lat/long using a button we provided.

All said, I’m pretty thrilled with the results and am excited to (a) use the widget to discover new places to go and (b) continue to enrich the dataset. In fact, it’s wide open for you to do the same! Mode makes it easy to collaborate so feel free to add more data, run your own analysis or let me know if something is missing (like can I bring my dog!?). Would also welcome any recommendations for Oregon and Washington must dos! 

Find me @SethTeicher if anything comes to mind. 

Explore the Data further