For years, advertisers have been able to target and exclude people using “dark ads.” Often, those ads have targeted and excluded minorities.
In 2018, the U.S. Department of Housing and Urban Development filed a complaint against Facebook, alleging that the social media platform had violated the Fair Housing Act. Facebook had done this, the complaint alleges, by allowing advertisers to discriminate against users based on sex, race, religion and even by interests such as “mobility scooter” and “deaf culture.”
“Facebook mines extensive user data and classifies its users based on protected characteristics,” the HUD complaint said. “Facebook’s ad targeting tools then invite advertisers to express unlawful preferences by suggesting discriminatory options, and Facebook effectuates the delivery of housing-related ads to certain users and not others based on those users’ actual or imputed protected traits.”
After the complaint, Facebook responded by saying that “there is no place for discrimination” and removed 5,000 ad target options. “While these options have been used in legitimate ways to reach people interested in a certain product or service, we think minimizing the risk of abuse is more important,” Facebook said in a blog post.
But the charges of discrimination by Facebook aren’t new; for the past three years, Facebook has been consistently hounded by charges that its platform allows for discriminatory targeting. A 2016 investigation by ProPublica found that Facebook advertisers could create housing ads allowing posters to exclude black people and, a year later, found that Facebook hadn’t fixed the issue. In 2018, the American Civil Liberties Union and three women filed a lawsuit against Facebook claiming that they were blocked from seeing job ads posted by 10 businesses that were using Facebook’s ad system to show job postings only to men—if true, this would likely be a violation of Title VII of the Civil Rights Act of 1964, which protects employees and job applicants from discrimination based on sex, race, religion and national origin.
These dark ads are made to be seen by specific groups but unviewable by others; some may refer to them as microtargeting or nano-targeting. The public became most aware of dark ads during the Cambridge Analytica controversy in 2018, when The Guardian reported that President Donald Trump’s 2016 campaign relied heavily on dark ads, often running 50,000 to 60,000 variations of Facebook advertisements each day. Brad Parscale, digital media director of Trump’s 2016 campaign and campaign manager for the 2020 campaign, told Bloomberg that they used these dark ads to target black voters. The aim of the ads was lowering voter turnout for Democratic candidate Hillary Clinton, Bloomberg reported.
Dark ads, especially those based on immutable characteristics such as sex and race, pose an interesting question to advertisers and marketers: When is it OK to exclude someone from seeing an ad? When is it OK to target based on an immutable characteristic, such as sex or race?
A New Opportunity
Guilherme Pires, co-author of Ethnic Marketing Theory, Practice and Entrepreneurship and professor of marketing at Australia’s Newcastle Business School, says that ethnic minority customers face several challenges in the market—bias, price discrimination, invisibility, language-related problems and perceived low critical mass.
“Invisibility is, in my view, a major source of concern because the implications for perceived critical mass detract from their being targeted,” Pires says. “In fact, even when critical mass is not the issue, a business may avoid targeting a specific group on account of negative reactions from other groups and, in particular, the mainstream.”
When used well, dark ads could overcome these visibility and critical mass problems, Pires says, as they allow marketers to target individuals with tailored value propositions. This can make dark ads relevant and beneficial for the targeted group while being inconspicuous to others.
“The possible ethics issue, in my view, has to do with the ‘how’ and ‘what’ is actually done,” Pires says. “Has permission been granted by the consumer to be targeted with dark ads? Has the content been abused, such as by containing pictures, endorsements or other content that is false, misleading or in some other way abused? Are there clear and easy ways for consumers to terminate undesired communications and get compensation for malpractice?”
These questions are relevant to all customers, Pires says. In Ethnic Marketing Theory, he and co-author John Stanton write that marketers must be especially aware of how people’s culture affects their behavior—and self-aware of how their own culture affects their behavior.
“The increasing visibility of minority ethnic groups … has heightened business awareness of the opportunities offered by ‘new’ markets based on multiple ethnicities within national borders,” Pires and Stanton write. “Yet, the perceived effectiveness of the increased multicultural marketing efforts to take advantage of those opportunities were estimated at only four of each 10 cases.”
Targeting by Exclusion
Targeting by group is ineffective, says Felipe Korzenny, author of Hispanic Marketing: A Cultural Perspective and professor emeritus of advertising, integrated marketing and management communication at Florida State University. But Korzenny says that targeting by exclusion, as had been allowed on Facebook, is clearly exclusionary.
“If you say, ‘I would like to reach everyone except for people in same-sex relationships,’ that would be exclusionary to sexual [orientation],” Korzenny says, as an example.
But Korzenny says that it would be OK to target by exclusion if the exclusion is based on a behavior rather than a trait. Even still, Korzenny says that targeting by exclusion or group will be much less effective than targeting by past behavior. “That’s what the big marketers do,” he says, citing Amazon as the best example of a company that targets consumers by how they behave. “They reach you believing that you will continue to behave in a certain way in the future and that is a lot more accurate than saying ‘Hispanic.’”
To further highlight the ineffectiveness of targeting by ethnicity rather than behavior, he notes that Hispanics share many common traits—the Spanish language, the Catholic religion, roots in Spain, an affinity for lime flavor—but if a marketer tries to sell lime soap, targeting people who have previously bought lime-scented items will cast a wider net than only targeting Hispanics. After all, there will be individuals who aren’t Hispanic who would love lime-scented soap and individuals who are Hispanic who detest lime-scented soap.
Another easy way to see why targeting by racial group can be ineffective: Imagine trying to target white people, a group that has come to encompass multiple ethnic backgrounds and live in all regions across the U.S. Black, Hispanic and Asian Americans have similarly expansive histories and backgrounds.
While there are some good ways to target by group—people who send money to Mexico are likely Mexican, Korzenny says, and people shopping for black haircare products are likely black—targeting or excluding by race, sex, sexual orientation or other immutable traits is largely ineffective and, at worst, unethical.
“You can almost always find a set of behaviors that would give you more bang for your buck because you might be able to include a bunch of other groups that do the same thing,” he says. While data-based behavioral targeting is likely more expensive than group targeting, Korzenny says that it’s also more precise.
Regarding exclusionary ads, Korzenny notes that—as the HUD complaint shows—it’s illegal to exclude people by traits they can’t change.
“That has been a gain by our society,” Korzenny says. “There’s still a lot of backward [thinking] going on. But most marketing departments probably have people who are a little more sophisticated than that.”
Editor’s note: On March 19, Facebook reached a settlement with civil rights groups and announced it will make “significant changes to its platform so that advertisers may no longer target or exclude based on characteristics like gender or race.”
For more information on marketing place inclusion and exclusion, see Journal of Public Policy & Marketing’s special issue.