,
12 tweets,
10 min read
Read on Twitter
Attention: If you are interested in #DarkPatterns (manipulative UX interfaces) or #Sludge (evil nudges), Jamie Luguri and I have an eye-opening new paper for you. (1/12)
papers.ssrn.com/sol3/papers.cf…
papers.ssrn.com/sol3/papers.cf…
As best we can tell it’s the first publicly available research to test experimentally how effective dark patterns are at manipulating consumers into signing up for services they do not want. (2/12)
In a lot of ways this paper is the next shoe to drop after @arunshemathur, @random_walker, @ineffablicious, @jonathanmayer & co-authors’ fabulous study demonstrating the prevalence of Dark Patterns (3/12)
@random_walker @ineffablicious @jonathanmayer Now Jamie and I are demonstrating something that companies have figured out but regulators haven’t. Not only are #darkpatterns becoming prevalent, but they are hugely effective in getting consumers to do what companies want them to do. (4/12)
@random_walker @ineffablicious @jonathanmayer A choice between “Yes please” and “Maybe later” for example, seems to hugely increase the percentage of consumers who will “consent” in comparison to a neutral “Yes” or “No” interface. Labeling one choice as “recommended” also has a big effect. (5/12)
@random_walker @ineffablicious @jonathanmayer How do we know? We ran an experiment on 1,963 American adults, told them we had signed them up for an identity theft protection plan and then randomly assigned them to one of three conditions to reject it: No dark patterns, mild dark patterns & aggressive dark patterns. (6/12)
@random_walker @ineffablicious @jonathanmayer The headline result: Whereas only 11% of subjects accepted the plan in no dark patterns condition, 26% accepted it after exposure to mild dark patterns and 42% accepted it after exposure to aggressive dark patterns. A small UX change caused big change in behavior. (7/12)
@random_walker @ineffablicious @jonathanmayer What’s more: less educated people were especially vulnerable to dark patterns. See pages 27-28 of papers.ssrn.com/sol3/papers.cf….
The distributional effects of dark patterns are very troubling. (8/12)
The distributional effects of dark patterns are very troubling. (8/12)
@random_walker @ineffablicious @jonathanmayer Use of aggressive dark patterns generated a major backlash. Many subjects were very steamed. Mild dark patterns generated little or no backlash. So mild dark patterns look to be fiendishly effective in bending consumers to companies’ will, and market forces won’t fix that. (9/12)
@random_walker @ineffablicious @jonathanmayer There’s a bunch of other interesting findings in the paper, along with our data-driven proposal to identify those dark patterns that are so successful at manipulating consumers that they ought to be unlawful. (10/12)
@random_walker @ineffablicious @jonathanmayer We provide a deep dive into the current U.S. law restricting dark patterns too. So if you are interested in UX design, contracts, consumer behavior, behavioral economics, or privacy, download the draft. Comments most welcome! (11/12) papers.ssrn.com/sol3/papers.cf…
@random_walker @ineffablicious @jonathanmayer Coda: It’s frustrating that @SSRN, which is hosting the paper, itself employs dark patterns to prompt people to create accounts there. But you don’t need to create an account to read the paper if you can convince captcha that you aren’t a machine. (12/12)
