Instagram may have been the most effective social media network for Russian spies in their effort to sway America's 2016 presidential election toward Donald Trump.
That's one of several new revelations in two reports commissioned by the US Senate's Intelligence Committee's investigation into Russian misinformation, released on Monday.
Others include the fact that the big tech firms – Facebook, Twitter and Google – went out of their way to hinder research into their failure to prevent online propaganda efforts; and that Russia focused its attention on two core groups in an effort to get Trump into the White House.
First, were African-Americans, who were shown tens of thousands of ads that were focused on encouraging them not to vote – by focusing on divisive social issues and basically saying "what's the point in voting?" Eighty-eight per cent of African-Americans voted for Trump rival Hillary Clinton.
Second were conservative voters, who were fed a daily stream of anti-immigrant content, reflecting what had become Trump's defining campaign promise – to deal harshly with immigration. Trump's eventual victory was in large part thanks to this segment of voters.
But interestingly, according to a report produced by New Knowledge, Columbia University and Canfield Research, for the Senate committee, the Russian campaign on Instagram – a service that is picture-based and doesn't have the same share features of other networks – ended up being the most interactive in its broad-ranging influence campaign.
The propagandists had 187 million interactions with their Instagram content, compared to just 77 million on Facebook and 73 million on Twitter, according to the report, which reviewed posts between 2015 and 2018 (those posts amounted to 10.4 million tweets, 1107 YouTube videos, 116,205 Instagram posts, and 61,483 unique Facebook posts).
It's hard to know, of course, how much influence those interactions actually had, or if they proved to be more or less persuasive than other posts, but it's a fair assumption to make that high levels of interaction point to greater influence. That's assuming the interaction analytics are correct: social networks are known for getting their numbers wrong.
In any case, we can imagine the Kremlin being happy simply sowing small seeds of doubt, upsetting discourse, radicalizing citizens, and dividing America with exploitative and controversial messaging, let alone nudging the electorate in the direct of Trump.
Don't forget: a lot of people voted for Donald because they disliked Hillary and the Washington DC establishment that much, regardless of whatever the Russians were pushing. Some folks are skeptical the social media memes had any impact at all. "It's far more likely that the Russians were just throwing a bunch of shit at the wall and seeing what stuck," said FiveThirtyEight's Nate Silver.
On the other hand, fewer than 100,000 citizens in three key states decided the outcome of the White House race, so it's not as though the Russians had to reach and persuade all 250,000,000 eligible voters to snub Clinton.
Notably, the report concludes that the Russian front – the Internet Research Agency – ran things like a modern online marketing campaign. It identified specific topics and target groups, and then set up presences on every main social media network – even Google+ – using coherent messaging and even design (fonts, colors etc) in posts that then linked to one another. One topic the researcher chose to highlight was the campaign's efforts to ride on the back of the Black Lives Matter movement with its "Black Matters" parallel campaign.
The Russian team also had an effective feedback system and would focus on specific messages and platforms that proved to be effective, using the same memes across their fake personas if they received a lot of interaction, and putting more effort into creating content for platforms like Instagram that appeared to be having a bigger impact.
Examples of the campaigns are exactly what you would expect in America: guns rights, police violence, demonizing Hillary Clinton, claims that Muslims were trying to introduce Sharia Law into the US, condemning immigrants, and so on.
What is also clear is that the campaign's main goal was to boost support for Donald Trump as president. Every message and post somehow tied back around to Trump, with even seemingly supportive messages to Democrats pushing courses of action that would benefit the reality TV star.
For example, supporters of Clinton's Democratic rival Bernie Sanders were a particular target and the sense that the Democratic party machine was working in favor of Clinton was used to persuade Sanders supporters – who would be unlikely to vote for Trump – to simply abstain from voting.
The second Senate-commissioned report, written by Oxford University's Internet Institute, reached the same conclusion: that the Russian campaign was large, sophisticated, and focused on Donald Trump's election as president.
Thanks, no thanks
In this report, however, researchers also take time to criticize the response of the social networking giants to their efforts to understand what had happened: the internet titans were extremely unhelpful, even after being publicly chastised in the press and in Congress.
The worst offender may have been Google, which supplied very little information and when it did, supplied in it hard-to-search PDFs, making it difficult and time-consuming to analyze. Facebook was no better: simply refusing to hand over information and limiting what it did send to English-language pages. Even the most responsive company – Twitter – only sent the researchers shortlinks, as opposed to full URLs, making it harder to use other tools to track their impact and links across the internet.
The New Knowledge report says the same, noting that the companies also appear to have stripped meta data from the information they sent i.e. they actively tried to disrupt efforts to understand the reach and impact of Russian propaganda efforts.
In short, the two reports tell us what we already knew: that there was a large, organized Russian campaign in favor of Donald Trump; that the campaign used divisive social issues to attract people's attention and push its messages; and the tech companies were caught completely unawares and then responded incredibly defensively when the size and scope of the propaganda campaign was revealed.
The difference from previous dossiers is that these reports are comprehensive and detailed. And they clearly identify the strategies and targets where previously much of the detail was anecdotal or intelligent conjecture. And, of course, we learned that Instagram punches above its weight, and the Russian campaign was so well resourced that it even bothered to post on Google+. ®