Identity is very important in the gay community. This need to belong has grown into countless labels, social groups, and cliques. To help understand and identify these social groups, I created the Gay Cliques Census and several interactive data visualizations to explain the results.
I started the survey in 2011, collecting initial data to help scope the project. Once I had enough statistically relevant information, I created a data visualization that relayed results live. Three years and 50,000+ survey responses later, I've created an infographic that segments the results by geographic region.
Collecting data on a community
I built a simple form to collect and store survey responses in a database. By using heat maps and form completion recording, I was able to remove questions that caused survey participants to abandon the form.
Optional fields like location are simplified with auto-completion, HTML5 geo-location, and fallbacks to a reverse-IP-location database.
Visualizing the results
Social groups were dynamically created from survey results. I released simple charts to illustrate these groups as statistically relevant data became available. Eventually I had enough to compile a live infographic.
The infographic includes a tool that lets people see how they compare with the survey results. People can then share their results or give negative feedback that helps refine the results of the survey.
Nearly 25,000 Likes, 700 tweets, 400 G+ Shares, and 1400 comments later, the interactive tool helped grow the survey through viral sharing.
Showing results by region
With enough data, I was able to visualize these social groups by geography.
When I filter survey responses by geographic region, all sorts of interesting things happen. We see what regions have the most passive groups (Santa Clara, California), where in the world Twinks are most popular (Canberra and New York), and more.
Just like the previous infographic, I've included a tool that lets people compare their stats with the results of the survey...only now with a geographic context. By answering a few simple questions, the tool can accurately determine where in the world you're considered attractive.
Keeping 20,000+ people updated and engaged
Nearly 50% of everyone who took the survey provided their email address to receive updates as the census progressed.
I created a series of autoresponders and drip campaigns to keep people informed, retain users, and encourage sharing of the census. ~80% of all survey responses came from social referral, and the more often someone came to view the infographics the more likely they would share it with their friends.
At the time of this writing, the email list has a 46.3% open rate.
Using this data to identify at-risk groups
While using census data to find where you belong or considered attractive is interesting, I was also collecting information about sexual partners and behaviour.
With enough data, its possible to cluster groups by sexual risk factor. Depending on their identity, certain groups were at much higher risk than others.
My intent is to use this data to help refine public health targeting for groups that need it the most.