What’s worked for social media for twenty years no longer works all of a sudden for Twitter. The software engine responsible for its growth - its internal algorithm for engaging users - is exacerbating declines in advertising revenue and is out of line with the company’s new vision. The opportunity is to align the algorithm with Twitter’s new vision to get advertisers back and lead social media into a new future, where products that profit from divisiveness transform into ones that thrive on coexistence.
When Elon Musk purchased Twitter just over a year ago, he opened the platform to people and entities previously banned or de-emphasized. Many of them were conservative. Soon there were headlines about rises in hate speech involving Black Americans, gay men, and Jews. As a result, advertisers started pulling their ad dollars because Twitter no longer felt like a safe place for their brands and products.
For many advertisers, this is a relatively easy decision. Twitter typically represents a small portion of an overall advertising budget. So if it feels risky, advertisers know they can easily redeploy the investment to other media, such as Facebook or TV. The impact of removing Twitter from an advertising plan is minimal to advertisers, but it’s devastating to Twitter.
Compounding the issue for Twitter is the misalignment of worldviews between marketers and changes to Twitter’s audience. According to our research using more than two billion dollars of political donation data from the FEC, marketers skew quite liberal at 68%. Twitter users also historically skew liberal. In 2022, seven of the top ten accounts for followers were either Democratic politicians or Hollywood celebrities (Elon Musk ranked #8, just behind Lady Gaga).
Now consider the fact that Twitter is attracting more conservative customers. According to Morning Consult, Democrats and Republicans now trust Twitter at about the same levels - 34%. Before Musk took over Twitter, 38% of Democrats trusted Twitter, while only 27% of Republicans did.
So marketers - the people holding the purse strings of advertising budgets - will be less comfortable with Twitter’s changes because they conflict with their worldview. If you’re selling something (Twitter advertising) to a liberal market (advertisers) and your product (Twitter) is becoming more appealing to conservative customers, then this causes problems with product-market fit. Brands and marketers will focus on the documented changes in hate, but they will also have a negative intuitive reaction to an increase in conservative customers.
According to reporting by The New York Times, Twitter’s ad revenue has fallen 59 percent since Musk took over. This is remarkable, given that 90% of Twitter’s revenue comes from advertising.
Musk is trying to solve this problem by hiring new Twitter CEO, Linda Yaccarino, the former head of advertising sales for NBCUniversal. She knows the largest advertisers in the world and how to sell to them, so it seems like a perfect fit. But it will take more than relationships to convince advertisers that Twitter is a safe place for their brands - it will take a transformation of what drives their business.
Elon Musk wants Twitter to become a more representative “global town square” where there is diversity in worldview as liberal and conservative. Linda Yaccarino echoes this desire in her early tweets, indicating that a global town square would “drive civilization forward through the unfiltered exchange of information and open dialogue about the things that matter most to us.” She also drops this vision bomb: “We have the opportunity to reach across aisles, create new partnerships, celebrate new voices, and build something together that can change the world.”
It’s a bold vision. The problem for Musk and Yaccarino is that decades of engineering are working against them, and that needs to be fixed. Those decades of engineering reflect a business strategy that is now out of sync.
The strategy for success in generating revenue on social media platforms over the past twenty years has been remarkably simple. Your goal is more ad dollars, so you need more impressions to sell more advertising. The best way to get more impressions is to amplify content with the most engagement in the form of likes, comments, and other actions - all managed by the internal algorithm.
Unfortunately, this kind of algorithm promotes divisive content because divisive content gets engagement. “Enragement gets engagement,” as Kara Swisher often says. But engagement is counter to getting advertising investment back for Twitter. It also doesn’t align with the vision of a global town square that “reaches across aisles.” So engagement from enragement for Twitter is a waste of time.
Musk and Yaccarino need a new algorithm aimed at something completely different that runs counter to more than twenty years of social media algorithm history: Moderateness. It’s the exact opposite of what social media algorithms focus on today. Yet it’s the only path for Twitter to realize its vision and regain its footing with advertisers.
On the face of it, an algorithm that amplifies moderateness seems like a non-starter from a business perspective. It’s not news if it’s moderate. Moderateness is often dull. It’s not entertaining to watch two people tolerate each other. Even sportscasters these days feel a need to yell into the TV camera for no apparent reason other than to get your attention.
The alternative to amplifying moderateness is more moderation - proactively filtering content to avoid problems. This leads to managing in circles. Content moderation is just quality control for the underlying problem created by an algorithm designed to amplify content that needs moderation. If you work in manufacturing and discover a high need for quality control, you don’t hire more QC people - you go after the problem.
In Twitter’s case, if the algorithm proactively amplifies moderateness, extreme content will make less money, and extreme content creators will find another venue. Or, if they are mercenary business people, they will pump out moderate content.
The big question is if something less “exciting” can succeed for Twitter. Your gut reaction is probably “No!” but here’s why it might work: Because it has to.
The financial incentives for Twitter and their stated vision line up with a strategy of moderateness. It will get ad revenue back and help realize their vision. And if there’s a billion dollars of will, there’s an army of software engineering way. Twitter can solve the algorithm issue, and once they do, the economic incentives of divisive voices feeding off the platform will change.
Moderateness - without all the moderation - will attract advertisers, especially when major brands like Bud Light, Target, and McDonald’s get swept up into culture wars on a moment’s notice. If they can turn the tide on advertising investment, then they have some time to figure out how to grow impressions with the new vision. You can imagine people like Elon Musk and Linda Yaccarino addressing the internal team, saying, “Make moderateness profitable by Tuesday or find another job” to get started.
There are two market forces at work here that also support a strategy of moderateness. First, most Americans are moderate regarding worldview as liberal or conservative, despite what each group believes about the other. Depending on the data source you look at - Pew, Gallup, or others - some 65% to 75% of American consumers are somewhat moderate. They are liberal or conservative or lean in either direction, but they are not at war. So the moderate market is huge.
Second, an interesting Pew study showed that 76% of Americans want both sides of a story covered equally, including the perspectives of liberals and conservatives. This contrasts with 44% of journalists who hold the same belief. So the market is not only large, it appears to desire to hear from both sides - but perhaps without the histrionics.
This strategy could fail for many reasons, but what if it worked? It’s hard to imagine, given the polarized environment we live in. But divisiveness from polarization is a relatively recent phenomenon, beginning in the 1950s and accelerating in only the last twenty years. Today’s polarization is just a blip in history, mostly driven by economic incentives that can shift when you point them in a different direction. Maybe that can start now for Twitter. Maybe.