Influence operations on social media that target politicians, voters and elections have proliferated in recent years. Russia’s hacking of the 2016 U.S. presidential campaign, and its flooding of fake news and stolen data into the public realm, is the most notable example of this trend.
But influence operations are also infecting the commercial space. Companies increasingly find themselves and their brands under assault from sources, often anonymous, who are trying to stealthily sway public opinion and consumer preferences against them. This manipulation of conversations and debates on Facebook, Twitter, Instagram and other social media platforms can impact a company’s bottom line, producing effects that range from annoying but largely neutral to serious and incredibly adverse. Companies need to consider the new ways in which they might be vulnerable to such operations.
Influence operations do not shy away from large targets. Nike recently came under digital attack—a coordinated, operational campaign, not solely angry rhetoric—after it rolled out a global advertisement in September 2018 featuring NFL quarterback-turned-civil rights activist Colin Kaepernick. According to a detailed technical study recently conducted by Morpheus Cyber Security and the global communications consultancy, APCO Worldwide, this cyber operation against Nike came from conservative and Republican political activists who were focused on driving down the shoe and apparel company’s sales and share price.
The Nike advertisements featured Kaepernick in both TV and print, saying, “Believe in something. Even if it means sacrificing everything. Just Do It”—an obvious reference to the controversy he first stirred in 2016 when he began kneeling during the national anthem at NFL games in order to protest inequality and police brutality.
The ad campaign re-ignited debate and embroiled Nike in the controversy. Nike’s share price fell 3.2 % the day after the campaign aired, and many users online called for a boycott of the company. Others filmed themselves destroying Nike products. On Twitter, hashtags appeared with slogans such as #NikeBoycott or #JustBurnIt. U.S. President Donald Trump himself assailed Nike in a September 5 tweet, a day after Nike launched the ad: “[…]Nike is getting absolutely killed with anger and boycotts,” President Trump wrote. “I wonder if they had any idea that it would be this way?”
However, a close study of Twitter traffic—conducted by Morpheus and APCO—illustrates that the consumer and political outrage expressed towards Nike wasn’t nearly as voluble or widespread as it initially appeared.
Indeed, much of the online debate was stoked by influence operations run by supporters of President Trump or other self-identified conservatives or Republicans (although, as usual with influence operations, state-level actors cannot be ruled out entirely). These activists created inauthentic Twitter surges, either by closely organizing echo chambers to mobilize tweets or by computer-generating traffic with bots. In the Morpheus-APCO study, an inauthentic source for Twitter traffic is called a nasnas, an Arabic term referencing a mythic devilish provocateur. Analyzing the Twitter traffic that appeared the first week after the advertisement aired, Morpheus and APCO identified a sample of 668 users who were attacking Nike for its Kaepernick campaign; each posted at least ten negative tweets or retweets about the company. These active users, however, didn’t maintain their campaign for long. By September 19, less than 10% were still vocal about the topic.
Further inspection of the active users revealed that 426 out of 668 sampled users were nasnases, or ones that exhibited inorganic traits—namely tweeting or retweeting more than 50 tweets a day, mostly about political issues. Comparing the number of total tweets an account has posted relative to the account creation-date can also reveal inorganic traits; for example, if an account was created in March 2018 but has posted hundreds of thousands of tweets, the account is likely a nasnas.
One of the online groups promoting the boycott against Nike was a website called TheNewMovement.org. Its founder, Tony Valenzuela (distinct from Tony E. Valenzuela, the award-winning director), openly identifies as a Trump supporter on the site and claims he switched from the Democrat to the Republican Party after saw the “hatred and lies thrown at [Trump] by the lonely left.”
TheNewMovement.org sought to organize an echo chamber of like-minded supporters to promote a ‘tweet pack’ of content that assailed Nike and the Kaepernick ad. Among them: “We stand with law enforcement while @nike stands with @kapernick7.” The website also suggested that users wait a few minutes after posting the first pack of ten tweets before posting the next pack of ten, probably to avoid getting suspended by Twitter.
This coordinated influence campaign by TheNewMovement.org was short and generated about 2133 tweets and retweets by 300 users. Fifty-eight of those users were also part of the sample group analyzed by Morpheus-APCO, as described above.
The conservative influence did not stop with TheNewMovement.org campaign. The study identified at least ten ‘leader accounts’ (e.g. accounts that other users frequently retweet), all displaying support to conservative and Republican views.
Furthermore, analyzing the connections between the core sample group of 668 actively negative Twitter users shows heavy connectivity and coordination among them. 93.5% of these users are each connected to at least five other users among the group, creating an echo chamber of accounts working together to generate impact on Twitter through coordinated means.
The campaign against Nike petered out after only two weeks. At this point, it became clear that the Kaepernick ad never really posed a serious challenge to Nike or its brand, despite President Trump’s exhortations or other online mobilization. It is a case of an influence operation that was largely unsuccessful.
Perhaps counter-intuitively, companies could leverage the attention generated by influence campaigns for their own benefit. In Nike’s case, the Kaepernick ad appeared to help the company . Nike’s share price quickly rebounded and rose to an all-time high by the end of September. The company also announced that it registered a 10% increase in sales during the summer quarter, and Vox reported that Nike experienced a $6 billion boost in overall value post-Labor Day. The joint Morpheus-APCO study shows that in the new digital economy, sentiment towards companies and brands aren’t always as they seem; social media conversation is not necessarily a bellwether. Like the activities in the political space, amorphous actors may attack companies for competitive,financial or even political reasons—and companies need to consider how such influence operations may affect their bottom line. However, technologies now exist that can expose the method and even the motive behind these operations. Not only that, new technologies can even counteract them.