The discussion among the panelists focused on criteria for identifying potentially dark patterns, including:
- How regulators evaluate a company’s advertising and marketing behavior,
- What reasonable standards exist, and
- Whether and how empirical evidence can be utilized to distinguish potentially disparate effects on consumers.
Introduction to Dark Patterns: Regulator Focus and Framework
Dr. Befurt first turned to Ms. Liebner for her definition of a dark pattern from the viewpoint of an FTC attorney. Ms. Liebner pointed to a recent FTC report to illustrate types of behaviors that represent a dark pattern, including:
- Design practices that trick or manipulate users into making choices that they would not otherwise make and that may cause harm, and
- Actions that take advantage of consumers’ cognitive biases – e.g., to steer their conduct – or that delay access to information they need to make fully informed decisions.
Ms. Liebner expanded on her examples of dark patterns by discussing the FTC’s recent $100 million settlement in FTC v. Vonage, in which FTC allegations focused on how difficult it can be for consumers to cancel a subscription. She explained that in FTC v. Vonage, the FTC alleged that Vonage allowed consumers to easily sign up online for their subscription telephone services but included various hurdles in the process of cancelling the service - e.g., by requiring consumers to speak to a live retention agent over the phone, obscuring the cancellation contact information on their website, creating redundant procedures in which consumers had to go through several agents, imposing long wait times before being able to speak to an agent, subjecting consumers to dropped calls and unanswered calls, and, as a contractual means, forcing high-dollar early termination fees.
Ms. Liebner then provided valuable insight into how she, as an FTC attorney, would distinguish between a company employing a true retention strategy and a dark pattern designed to retain customers by making cancellation difficult. In this context, Ms. Liebner explained that any type of marketing practice is subject to the factors under the laws enforced by the FTC, such as:
- Whether there is a representation, omission, or practice that is likely to mislead consumers who are acting reasonably under the circumstances
- Whether an action causes or is likely to cause substantial injury that consumers could not reasonably avoid (and the injury is not outweighed by countervailing benefits to consumers or competition)
As an example, Ms. Liebner pointed to injury in the form of lost time, which played a role in FTC v. Vonage, FTC v. Amazon, and FTC v. Credit Karma. She also noted the FTC’s desire for consumers to experience simple cancellation mechanisms.
Segments of Consumers: Informed Choice and Trade-offs
The conversation then turned to focus on the consumer, as Drs. Befurt and Pauwels discussed dark patterns through the lens of a market researcher. Dr. Pauwels noted that much of his research centers on privacy preferences and specifically how consumers trade privacy for time. Certain segments of consumers may have different preferences; Dr. Pauwels suggested that as long as information is communicated in a transparent manner (i.e., consumers are allowed to make their own trade-offs), the use of certain patterns may be acceptable.
Mr. Kshirsagar noted that problems may arise when firms are concerned about retaining their customers, and as such, firms may take advantage of shifting preferences in trade-offs. He pointed to the FTC v. Vonage case as an example in which the trade-off was making a phone call (i.e., time) in exchange for cancelation. Ideally, firms in a competitive market would attempt to match consumer preferences; dark patterns come into effect when there is a mismatch.
Mr. Kshirsagar further explained that while the FTC has the “reasonable consumer” perspective, many states instead have the “credulous customer” standard. In certain states with this standard, the particular kind of customer matters. Mr. Kshirsagar gave the example of new parents who are stressed because of the obstacles involved in ordering a new diaper service, and, as a result, may require certain treatment that accounts for their specific challenges and circumstances. Ms. Liebner agreed, noting that target audience matters and that what is considered “reasonable” may depend on what group is being advertised to or targeted.
Identifying a Dark Pattern
The panelists then discussed the viewpoint of firms, contrasting firms that attempt to optimize to certain subsegments of consumers with those that potentially employ a dark pattern. Dr. Pauwels acknowledged that while companies aim to optimize profits, well-informed companies value customer retention and lifetime value. He noted that there is an opportunity for companies to determine the different preferences of consumer populations in order to offer more choices in trade-offs between services and personal data. Mr. Kshirsagar continued that in the long run, he believes a core challenge will be to determine the kinds of industries that are more susceptible to dark patterns. He cited pressure tactics (e.g., fake timers or number of available hotel rooms) in industries where consumers are acquired quickly, as well as industries where there is a long-term consumer value (e.g., subscribers to a newspaper) that may make cancellation difficult. Ms. Liebner noted that when deciding which dark patterns to litigate, the FTC focuses on types of practices rather than specific industries or companies. The panelists continued to compare companies’ advertising practices, with Ms. Liebner noting that subscription traps in particular can be damaging to the consumer and competition.
Ms. Liebner and Mr. Kshirsagar then discussed an approach to identifying dark pattern conduct, drawing from their experiences at the FTC and New York Attorney General’s Office, respectively. Ms. Liebner explained that targets may come to the FTC from sources such as consumer complaints, news reports, or referrals from other federal or state agencies, and the FTC may choose to move forward in litigating due to factors such as egregiousness of conduct, scope of harm to consumers, and ability to get meaningful relief for victims. Mr. Kshirsagar agreed and added that the New York Attorney General’s Office is also concerned with how to identify dark pattern conduct and find reports or data about those practices when there is a lack of consumer complaints. He cited his recent article and noted that the FTC and state attorney general’s offices are working to develop investigative techniques to collect data on the state of a market in order to identify industry outliers that may be employing dark patterns.
Dr. Befurt then posed an example of a consumer who is aware of an arguably “hidden” fee but chooses to continually purchase a product and asked the panelists whether consumers with awareness are still damaged. Dr. Pauwels, Ms. Liebner, and Mr. Kshirsagar all agreed that these consumers would be damaged, citing factors such as the importance of regulation and comparing the scenario to bait-and-switch and bundled pricing cases. Mr. Kshirsagar added that structures that are opaque to consumers create a detrimental effect on the market due to a lack of available high-quality information.
Potential for Empirical Evidence
The conversation then turned to materiality in dark patterns cases and the usefulness of empirical evidence. Ms. Liebner explained that an FTC litigation may be centered on a small amount of harm to a large number of consumers or a large amount of harm to a small number of consumers. When asked about the data that could be used in these cases, Ms. Liebner explained that it may come down to what data are available from the company, such as internal data if the company has done testing, or copy testing or consumer surveys if internal data are not available.
Dr. Pauwels highlighted the difference between stated and revealed preferences, warning that individuals may overstate their preferences for privacy in a survey context. He recommended conjoint analysis as a potential tool to examine how consumers may trade privacy for convenience. Mr. Kshirsagar then cited a recent FTC report that examined opt-out rates across internet service providers as an example of helpful learnings from empirical data. Drs. Befurt and Pauwels agreed that it can be difficult to ascertain true preferences and that different methods may be suitable for different purposes, such as A/B or copy tests in addition to other survey methods. Mr. Kshirsagar continued that cross-market data collection can be a useful method, citing an example of an investigation into internet service providers and quality of traffic to gauge whether certain internet service providers put Netflix’s internet traffic at a disadvantage.
The panelists closed with a discussion on the opportunity for industries and regulators to build trust through a focus on transparency with the consumer.