The folks at StinkDigital produced this nifty little digital execution.
The product is for “Honest Tea,” so they conducted a social experiment across the country to determine who was most honest. If they left tea out unattended, but then asked for people to leave a dollar for each bottle they took. Once they harvested the data, they created a slick data viz experience with a bunch of filters for adhoc comparisons.
As you would expect, the data visualization and over all UX experience are slick, but that’s not the interesting part to me. What is interesting is the way they handled the filters (e.g., “bikers” or “parents with kids”). You know they couldn’t create the filters by having folks fill out segmentation surveys, because that would kill the experiment. What was clever is the locations of the experiment fueled the assumptions for the filters. For example, Houston’s location was a zoo, so besides a geo filter, they likely made the assumption that anyone who interacts with the experiment is a parent with child.