Have you ever had that feeling that your browser is watching you? A product you looked at on Amazon is suddenly an advertisement on Facebook. Some cloud-based algorithm is cataloging your every move and predicting your ad preferences, but you never really see it.
Floodwatch is a means to fight back. A project led by data viz guru Jer Thorp, and his Office of Creative Research (OCR), Floodwatch is an extension for the Chrome browser that will catalog every ad that appears on your screen into a sleek, searchable, and sortable database for you to explore. And, if you like, it will share your results, so researchers might reverse engineer the secret algorithms that are profiling you and feeding you ads.
"I’m not against advertising, nor am I specifically against targeted advertising. But I do believe that people should be able to see what is being presented to them, so that they can get a sense of what kind of things advertisers know about them," Thorp tells Co.Design. "Really, the sticky point for me is that a large percentage of the data that is being used to ‘customize’ is not being given voluntarily. Personally, I’m really not okay with that, and I believe a lot of people would complain about these practices if they knew more about them."
After installing Floodwatch, you may never notice it again. But at any time, you can click its icon to be taken into a Floodwatch dashboard, which is an interactive visualization displaying every ad you’ve been fed in a timeline or in a sortable grid. In the grid, you can look at the actual ads that have appeared in your browser, organized by publisher (NYT or Fast Company), theme (jewelry, apparel, or banking), or even color. Why color?
"It’s a really nice effect, and I also think there is something encoded in the color set that is within your ad profile," Thorp writes. "I suspect, for example, that young women are seeing ads with very different palettes than young men are seeing."
It’s these ad profiles—the stereotypical sketches that ad networks have drawn of each of us—that Thorp’s team is really trying to deconstruct. After all, a targeted ad is more than just a targeted ad. It’s a series of assumptions made about someone, followed immediately by a collection of opportunities (ads that say "buy this" or "go here") they’re either offered or not. To OCR, being shown different ads based upon demographics like age, gender, income, or geography may actually be classifiable as discriminatory practice, and they’d like to build a dataset that’s large enough to test such a theory.
In the meantime, Thorp performed an interesting experiment for his own Floodwatch data, inspired by a project called Cookie Jar made by his then-student Julia Irwin. He paid 10 people $5 to profile him entirely based upon the ads he’d been shown, and he was kind enough to share the full texts with us, excerpted below.
As you glance at the results, keep in mind Thorp's actual background: he's a Brooklynite data researcher who pioneered interactive content at NYT before founding his own company to pursue other big projects. His work has appeared in the MoMA.
While some were clever enough to piece together that he was a creative and a Canadian, most of the results were either sad or hilarious, depending how you view them.
"The random individual who viewed web-based ads over a two week period is most likely an unemployed man in his early thirties. His affinity for employment ads gives credence to the fact that he probably does not have a job and hasn't had a job for a considerable amount of time…"
"...he must live in or around Los Angeles. Since he lives in that general area and also has ads for Vscore, he must be trying to break into acting. He probably moved from Missouri to Los Angeles in order to land some gigs...He could have went to University there and decided he'd rather be an actor. Either way, he has acting as his major career goal. He is not really getting anywhere with the acting so he looked around at other careers he could have while also doing acting on the side."
"John is his name. Perhaps not in real life, but in my re-creation of him...John wants the newest technologies, best jobs, most money, and status. Unfortunately for him, that takes work. While reading about Hollywood and dreaming of a better future, he is neglecting to make the most of his current situation. He needs to stop the dreaming and get down to business before life passes him by. There is no inclination he won't be in the same situation tomorrow, and the next year, and perhaps the next decade. Although a dreamer with hopes and aspirations of grandeur, John needs to focus on his current situation and build his future on reality, not dreams."
"I would say this person, in my opinion, is of the more elderly age group — probably around 65-75. They are most likely either retired or just about to be, and definitely male, based on all of the manly, rugged clothing ads...Honestly, I feel that any woman would be lucky to have this gentleman, as he seems to be fun, exciting and kind with a bit of an extravagant lifestyle... just a little lonely is all."
Floodwatch is available now. You can install it here, and it’s not necessary that you share your data with researchers to do so.
This post has been updated to include contributions by Julia Irwin regarding the promotion and testing of Floodwatch.