Microtargeting is a powerful advertising tool that allows politicians such as President Donald Trump to aim ads at specific groups of people, or even individuals, online. The advertising tech has become increasingly controversial, with critics saying it permits politicians to target very narrow groups of voters with tailored messages that have the potential to manipulate the political debate.
Now, two major tech companies, Google and Twitter, have overhauled their political ads policies and severely limited campaigns’ ability to target voters — and Facebook is under mounting pressure to do the same.
It all fits into an even bigger debate over whether restricting political ads online could do enough to curtail the type of election interference and misinformation campaigns that played out during the 2016 presidential election. Limiting how narrowly a campaign can target voters, or what kinds of ads politicians can pay for, may solve a few problems, but some say tech companies’ real concern should be moderating the misleading and extremist organic (that means unpaid) content that clogs their platforms.
Here’s the state of play: In October, Twitter CEO Jack Dorsey announced that his company would ban political ads from its platform. However, it will still allow issue ads, which means that while Elizabeth Warren can’t advertise her bid for the presidency, a political group can still buy ads urging people to vote to protect the environment. Twitter is limiting microtargeting for these issue ads, which will allow advertisers to target at the state level in the US, but they can’t use zip codes, and they can’t use keywords or interest targeting around politics (such as terms like conservative or liberal) to find their targets.
Twitter CEO Jack Dorsey arrives for a House Committee on Energy and Commerce hearing about Twitter’s transparency and accountability, on September 5, 2018.
Drew Angerer/Getty Images
Then last week, Google said it would restrict advertisers’ ability to microtarget political ads across its products, such as Search and YouTube. As of January 6, political advertisers will still be able to target people using age, gender, and zip code, but they won’t be able to get as granular with specific voters and groups as they used to.
Twitter and Google’s moves are especially significant because they put pressure on the platform that kicked off much of the recent debate around political advertising in the first place: Facebook. The company came under fire this fall over its policy of allowing politicians to lie in ads. Initially, CEO Mark Zuckerberg took a hard line on the matter and said Facebook wouldn’t back down. Now, weeks later, the company is in fact considering making some changes to its policy, including how much microtargeting it allows.
The changes (or in Facebook’s case, potential changes) have gotten a mixed response from politicians, advertisers, pundits, and the public. Given all the chatter about political ads, Twitter and Google doing something has been generally well received by voters. But there’s hardly a consensus on whether these are the right moves, or whether we’re even having the right conversation.
Microtargeting and how political campaigns use it to advertise, briefly explained
Twitter’s decision, while sure to stir up some controversy, isn’t a huge deal for politicians or for Twitter — it’s a relatively small player within the political online advertising space, and political ads don’t make up a big portion of its revenue. Google and Facebook, however, are much bigger deals, so I’ll focus on those.
One of the most important features Google is nixing is “Customer Match,” which allows campaigns to match people’s online profiles with the voter data they have and target them specifically. On Facebook, a similar tool is called “Custom Audiences.”
Here’s how it works: Political campaigns create databases about voters that include information about whether a person is registered to vote, how often they vote, their party affiliation, their mailing address, their email address, and their phone number. They can then upload those voter files to Google and Facebook to find those people’s online profiles, and then advertise specifically to them. Google is now banning political campaigns from doing this. So politicians will no longer be able to serve an ad to me, Emily Stewart, specifically. But they can still advertise to women like me, ages 25-35, in Brooklyn, New York.
“If a campaign knows where the people they’re trying to target live, then they can still target them on Google,” said Alan Rosenblatt, a director at progressive strategy group Unfiltered.Media. “That said, it isn’t as precise, and as a result, you’ll have to buy more Google ads to reach the people you want as frequently as you want.”
Google’s move puts further pressure on Facebook to change something when it comes to its political ads policies, but it’s not clear what the company will decide. Facebook’s vice president of global marketing solutions, Carolyn Everson, told Recode’s Peter Kafka last week that the company was not considering targeting — but then quickly walked that back. The Wall Street Journal later reported that Facebook is considering increasing the minimum number of people a political ad can target on its platform from 100 to a few thousand — but at this point, Facebook still hasn’t made any changes, so anything could be on the table.
Some strategists and politicians think this is great
Reactions to Google’s microtargeting decision and speculation Facebook might soon follow suit have been all over the place.
Proponents of limiting campaign microtargeting say this is a good way to root out some of the bad actors and bad practices in online political ads. It will make it harder to target negative and fake ads so precisely, so if a shady ad is out there, more people are likelier to see it, and that means political opponents and other groups will have a better ability to counter it. For example, in 2016, the Trump campaign targeted infrequent black voters with ads showing Hillary Clinton in 1996 calling some young black males “super predators.” It was a voter suppression effort aimed at getting those voters to stay home, and limiting microtargeting would make such an effort more difficult to execute in such a precise way.
Proponents of Google’s move also note that voters often don’t know they’re being targeted with an ad campaign. In a statement following Google’s announcement about ad targeting, Sen. Ron Wyden (D-OR) noted that the transparency features Facebook has around advertisements don’t apply to its custom audiences feature, for example. “Targeted influence campaigns are more effective and more cost-effective than blanket propaganda, and far harder to identify and expose,” he said.
Ellen Weintraub, chair of the Federal Election Commission, in a series of tweets applauded Google’s decision and challenged Facebook to follow it. However, she warned that zip code-level targeting might still be too narrow. “The public may not be able to adequately hear and respond to communication targeted down to that geographic level,” she said.
*Very* big news last night in internet political ads. @Google’s plan to eliminate #microtargeting is a move that – if done right – could help make internet political advertising a force that informs and inspires us, rather than isolating and inflaming us.
— Ellen L Weintraub (@EllenLWeintraub) November 21, 2019
Chris Nolan, the founder of political advertising and analytics agency Spot-On, told Recode that Google’s decision “changes the conversation about political advertising” in that it acknowledges the velocity of online advertising and how precise it can be in comparison to broadcast ads. “The largest player in the ad technology business has said that political ads shouldn’t be treated the same as brand advertising, and that is a big deal,” she said. “But the question remains, are other players going to follow along?”
And those other players are not just Facebook, but also other platforms that facilitate digital ads, such as SpotX and Adobe.
Others are not so happy with this
Of course, not everyone has been thrilled with Google limiting microtargeting — or the prospect of Facebook doing so. They argue that focusing on microtargeting is misguided, and that it could undermine up-and-coming campaigns that can’t afford television ads but that can run online ads with a small budget.
Tara McGowan, the founder of Acronym, a progressive strategy group that is currently raising $75 million to counter the Trump campaign’s digital ads, said in a statement to Recode that Google’s decision won’t curb disinformation but instead will “hinder campaigns and others who are already working against the tide of bad actors to reach voters with facts.” She said the decision affects Google’s ad inventory as well as inventory across the internet. “They are essentially using their market power to limit how campaigns can speak to voters where they get their information,” she said.
Progressive digital network Blue Digital Exchange warned in a Medium post signed by dozens of operatives and strategists that limiting microtargeting will ultimately harm Democrats more than it does Republicans because the Republican voter base is more homogenous and therefore easier to reach with limited targeting. “Without using voter registration files, a Republican campaign can advertise to this homogenous group of people and expect to engage a majority of registered voters. In contrast, a Democratic campaign advertising to young, urban, people of color might have to spend a lot more because they are trying to reach a fraction of users within that group who are actually eligible to vote,” they wrote.
But Trump’s campaign is also displeased with Google’s decision. His campaign manager, Brad Parscale, slammed the policy change, and Trump’s campaign and other Republican leadership groups put out a joint statement saying the changes Google is making are an attempt to suppress voter turnout. (There’s no evidence that is true.)
The Trump camp has also been sending out warnings to voters about Facebook potentially limiting microtargeting. It’s true that limiting microtargeting would probably slow down the advantage that a big, well-funded campaign has in placing hundreds of microtargeted ads that are impossible to track. But it’s also true that those in Trump’s orbit are eager to jump at any opportunity to take a swipe at big tech companies and accuse them of anti-conservative bias.
Political elites & Big Tech want to rig elections – Dem primary & 2020 included. They’re targeting Trump because he’s the big dog, but they’re also after Dems like Sanders & Warren.
Won’t stop until they control all digital political speech.https://t.co/kpDpIAmHV7
— Brad Parscale (@parscale) November 21, 2019
Jesse Blumenthal, vice president of technology and innovation at Stand Together, a network of Koch-funded nonprofits, told Recode that beyond the ins and outs of microtargeting policies, it’s a question of free speech. “Implicitly, what is being said is political speech is less valuable than other forms of speech and that these companies ought to be the ones moderating political speech,” he said.
Of course, it is the government that owes Americans free speech and is bound by the First Amendment not to infringe on their rights. The law doesn’t apply in the same way to Twitter, Facebook, and Google, so they can police their platforms however they want.
Google, Facebook, and Twitter would rather talk about political ads than a lot of other issues on their platforms
All the chatter about political advertising and microtargeting distracts from much bigger and more complicated problems for tech companies to deal with: namely, the organic content that often spreads disinformation and inflammatory, violent, and abusive content across their platforms.
When Google is talking about political microtargeting, what it’s not talking about is how YouTube’s recommended and autoplay features push people toward more radicalized and extremist content. While we’re debating whether or not Facebook should permit politicians to lie in ads, what we’re not discussing is how fake news, memes, and disinformation spread organically on the platform. Twitter is banning political ads, but it has a much bigger challenge in addressing how the platform is used to spread hate, the president’s use of his Twitter account, and the abuse and harassment that happens on its platform every day.
The real misinfo and political abuse that happens on social is organic. YouTube discovery, Twitter bots, Facebook memes (groups!!!!), etc.
Ads are a fraction of the problem. When a tech company “fixes” ads without mentioning what they’re doing on organic, they’re playing you.
— Rob Flaherty (@Rob_Flaherty) November 21, 2019
“Instead of monitoring and taking responsibility for the spread of misinformation on their platforms, Google has chosen to pursue a disingenuous and frankly dangerous shift in their policies so they can claim publicly to be serious about the problem,” McGowan told Recode.
There’s still a lot of anxiety about the 2016 election and the roles of disinformation, foreign interference, and tech on politics and democracy around the world. Microtargeting has always existed (just look at mailed advertising campaigns), but obviously the internet amplifies it. Focusing on digital political ads is a way to address one problem, but it’s not the same as dealing with the broader question of the role of Big Tech in shaping the future of political discourse and democracy.
Vox · by Emily Stewart · November 27, 2019