loader image

Contagion: The Spread of the Coronavirus

Over the last 28 days, the most successful 25 global page posts on Facebook featuring the word “coronavirus” have created 5.7 million total interactions, 3.4 million shares, and have generated over 216,000 comments from Facebook users.

The top post of all, published by CNN, resulted in 1.3 million interactions alone, was shared 891,000 times and was replied to with 57,000 comments.

These already seem like big numbers, however the video contained within the post has been viewed 96,941,410 times at the time of writing – reaching 14 million more people than the 2019 population of Germany with one single social media post.

On only a cursory search, we can identify 50 global Facebook pages who have mentioned “coronavirus” in at least one post in the last 24 hours alone. These 50 pages, mostly news organisations, have a combined audience in the hundreds of millions and generate between 3 and 200 posts per day, each.

It is not difficult to see how a single news event – or any information or disinformation relating to it – can travel so far so quickly, and this is only Facebook. With information shared across other platforms including Whatsapp, Snapchat, and TikTok the entire global population is accessible to any actor at the touch of a few buttons.

Taking Twitter, for example, we can pinpoint the first mentions of the coronavirus in the last four weeks as occurring on the 31st of December 2019, when the term was used 126 times. Mentions then began to build, holding steady around the 1,000 posts mark before exploding on the 20th of January 2020 – peaking at 201,136 tweets at midday (UCT) on January 26th.

To give a rough estimate of reach (the number of people who could have seen the tweets on January 26th, sometimes referred to as “impressions”) we can apply a simple industry calculation.

The average number of Twitter followers per user is currently 707. All we need to do is multiply the number of tweets by that number of followers. We can estimate that potentially 142,203,152 saw the tweets during the peak on January 26th.

Over the whole four week period 4,772,200 tweets were generated featuring the word coronavirus, meaning the content was potentially seen 3,373,945,400 times (or seen 10 times each by the 330 million active monthly users of Twitter).

And that is without considering how many of those tweets have gone on to be embedded in news articles shared in other ways across other platforms, websites, and by email.

If you consider that over 80% of the world’s leaders are Twitter users, there is little doubt they influence and are influenced by platform content.

Following a long term trend of where Twitter content is most created in response to media events, 22.8% of the 4.7million tweets were geo-coded to the US, despite only 22% of Americans being on Twitter according to official platform statistics.

This means that Americans have been somewhat disproportionately exposed to coronavirus content on Twitter, despite a handful of cases having been confirmed by the CDC.

Paris, New York, London, and Hong Kong have seen equally high percentages of the content.

We can identify that the majority of Twitter content has been shared in the majority by Android users – if used as a demographic data point telling us the income bracket is lower to mid-working class.

We can also identify the words which have dominated in terms of repeated inclusion in tweets featuring the word coronavirus.

The elevated use of Spanish words is confirmed within further data which shows us that 57% of the total tweets were logged in English, while 17% were in Spanish. This number then reduced to 8% for French and then down under 5% for all others.

This identification of languages used and the confirmation in the words used is particularly relevant when assessing whether or not known disinformation actors have been active within the topic.

We can immediately see that RT’s Spanish language arm has been one of the major actors in the topic and is the leading Spanish language source of shared content discussing the coronavirus.

Taking one of the 32 Spanish language links to an RT video article, the sharing history of the link is visible.

In this case, the link was posted 24 times across Facebook, Twitter, Instagram, and Reddit, to an audience of 8.1million followers and generating 5,539 interactions.

Turning away from social media and into web traffic, we can identify that RT’s Spanish language service performed exceptionally well during the same four week period, averaging well over 100,000 daily visits in the US and over 300,000 daily visits in Mexico, and over 200,000 daily visits in Spain.

It is apparent that the coronavirus topic became viral in a relatively short space of time and content generated has been sufficiently spread by a number of different methods to ensure that every corner of the world has at least seen some content relating to it, whether by website, Facebook, Twitter, or elsewhere.

In respect of which spreads quicker, the virus itself or the digital content relating to it, the answer is that information still travels faster than any pandemic. The only difference being there is no chance whatsoever of containing the digital content.

Related Articles

Losing Narratives

Narrative isn’t a toy, you see. Yet it’s played with all too often by people who don’t understand it. Narrative is a weapon in the hybrid warfare arsenal and has tendency of backfiring if not handled with care.