The Timeline: February to July 2022
This study zeroes in on a six-month stretch of the internet's chaotic response to geopolitical crisis, from February 1 to July 31, 2022. Russia invaded Ukraine on February 24.
Within days, hashtags like #IStandWithRussia and #IStandWithPutin began to trend—not organically, but as part of what now appears to be a coordinated propaganda campaign.
The researchers collected 349,455 pro-Russian tweets and mapped how these messages spread. The result?
A sobering digital autopsy of an influence operation that reached an estimated 14.4 million users, with bots doing much of the heavy lifting.
Anatomy of a Hashtag Campaign
The campaign was hardly subtle.
Four hashtags served as the backbone of the pro-Russian message machine: #IStandWithRussia, #StandWithRussia, #IStandWithPutin, and #StandWithPutin. These weren’t just casual posts from rogue accounts.
On March 2 and 3—right as the United Nations voted on a resolution condemning Russia’s invasion—the volume of tweets spiked dramatically, with over 100,000 pro-Russian messages posted on March 3 alone.
The coordination wasn’t just about volume. It was targeted. Countries like India, South Africa, and Pakistan—all of which abstained from the UN vote—were flooded with these tweets.
On the day of the vote, 41.7% of all messages came from accounts geolocated in India. Let that sink in.
Enter the Bots
A postmortem wouldn't be complete without identifying the agents of dissemination.
About 20.3% of accounts spreading pro-Russian content were bots, according to Botometer, a tool that analyzes Twitter accounts for bot-like behavior.
Many of these bots were created in late February 2022, coinciding with the start of the invasion.
And they weren’t just there to shout into the void. Bots were strategic amplifiers.
They retweeted human-generated content to inflate the apparent popularity of pro-Russian narratives.
By generating artificial engagement, these accounts helped push pro-Russian hashtags into Twitter’s trending topics, increasing visibility to unsuspecting users who might assume popularity equals legitimacy.
This strategy marks a departure from previous bot playbooks. In the 2016 U.S. election, for instance, bots targeted influential users hoping to provoke retweets and reach.
Not here. In this campaign, bots went for breadth, not depth.
They weren’t trying to persuade elites—they were trying to flood the zone.
Bots as Hype Men, Not Authors
One of the more revealing insights from the study is that most of the content came from humans.
Only about 21% of original tweets were from bots, but bots accounted for over 25% of the retweets.
Think of them as the online equivalent of studio audience laughter: they're not telling the jokes, but they're making sure you hear the punchlines.
Interestingly, these bots weren’t retweeting each other. Most retweeted humans, particularly in the early stages of a tweet's life cycle.
That early amplification helped push messages past Twitter’s visibility thresholds, nudging them closer to viral status.
Whose Voices Were Loudest?
Compared to their pro-Ukrainian counterparts, pro-Russian accounts had fewer followers and lower engagement.
The average pro-Russian account had 1,690 followers; the average pro-Ukrainian one had 2,248. Pro-Russian tweets also earned fewer likes, replies, and retweets on average.
And yet, through sheer volume and coordination, they managed to create the appearance of a movement.
It was less about persuasion and more about perception. A digital sleight-of-hand.
Geopolitics Meets Platform Dynamics
The paper’s most damning implication is that Twitter (now X) served as a fertile ground for this operation.
Despite its content moderation and bot-detection efforts at the time, the platform allowed a sizable propaganda campaign to unfold in real time.
Moreover, the campaign was tailored to exploit geopolitical ambivalence.
Countries with significant English-speaking populations and less entrenched positions in the NATO-Russia binary were primary targets. India, South Africa, and Pakistan were flooded with content designed to push anti-Western, pro-Russian sentiment, often by exploiting historical grievances or anti-imperialist rhetoric.
The Politics of Visibility
One particularly alarming tactic involved bots repeatedly engaging with content to game Twitter's trending algorithm. The goal?
Get hashtags like #IStandWithPutin in front of as many eyes as possible. This strategy played on a basic truth of social media: visibility can equal legitimacy, especially for casual or undecided users.
Repeated exposure is also psychologically potent.
The study references the "illusory truth effect" — the idea that people are more likely to believe something if they've seen it multiple times.
In this context, seeing a hashtag trend, even if artificially, can influence beliefs and undermine support for Ukraine.
A Tale of Two Networks
The study includes an analysis of retweet networks that paints a picture of echo chambers tailored by geography.
In India and South Africa, accounts retweeted each other in tightly knit clusters. In contrast, U.S.-based accounts were scattered, less coordinated.
That kind of localized echo chamber suggests not just random support, but orchestrated efforts to entrench pro-Russian sentiment in specific national contexts.
A Propaganda Campaign with Limits
Despite its scale, the campaign had limits.
Pro-Ukrainian content vastly outnumbered pro-Russian tweets overall. And bots promoting pro-Russian narratives were largely ineffective at getting humans to retweet them.
This isn’t a case of a successful manipulation as much as a failed attempt that still managed to make noise.
But even a failed attempt teaches us something: that the infrastructure for influence operations remains disturbingly intact.
And that should be enough to keep policymakers, platforms, and the public awake at night.
Lessons from the Postmortem
What does this digital autopsy tell us?
First, that social media remains dangerously susceptible to manipulation. Second, that bot armies don’t need to be persuasive to be effective—they just need to be loud, fast, and well-timed.
And third, that we can’t rely on anecdotal detection.
It takes large-scale empirical research like this to uncover the full scope of propaganda efforts.
As the authors conclude, strategies to mitigate these campaigns need to focus on curbing bot activity—particularly during geopolitical flashpoints when the stakes are highest.
We may not be able to prevent the next propaganda campaign from launching.
But with studies like this one, we can at least see it coming.
Citation: Geissler, D., Bär, D., Pröllochs, N., & Feuerriegel, S. (2023). Russian propaganda on social media during the 2022 invasion of Ukraine. arXiv:2211.04154
