Exploring South African Hashtag Virality

Web Application

Below is my exam project I made for my university course Introduction to the World Wide Web. I tracked and analysing South African meme virality using a web application with the Google Maps Embed API and some graphing using desmos. I hope you find reading it as interesting as I found making it.

To choose which hashtag you would like to view on the map, click the button in the top left corner of the map. There you will also find the legend.

  • #WooliesWaterChallenge started when a group of young men posted a video of themselves drinking Woolworths water, which prompted them to change their accent after taking a sip. Many people created their own versions of the original video, including Mmusi Maimane, the Leader of South Africa’s opposition Democratic Alliance (DA) political party. The original video now has more than 6,500 retweets and 39,000 views on Twitter.
  • #HotCrossBuzz started when a video of a police officer eating a hot cross bun and taking a breathalyzer test reading 0.21mg of alcohol per 1 000 ml of breath. The legal limit is 0.24mg, meaning that a few bites of the bread has nearly made the officer “too drunk to drive”. This video sparked great controversy in the ingredients of the buns, reaching more than 5500 retweets on Twitter.
  • #MyVokMarelize started when a video surfaced of a young girl crashing into a rugby post on a bicycle and as she falls to the ground, her mother, who was recording the bicycle riding lesson, exclaims in Afrikaans, “my f** Marelize”, as she begins to walk towards the girl. The gained much attention from South Africans as a comical reaction of amazement, disbelief and amusement. The original video gained over 450 retweets on Twitter.



The internet is a powerful new space, allowing for the sharing of ideas with millions of people across the world within seconds. It allows people to research, communicate and relax. There is such an abundance of information on the internet, that the modern problem is not lack of access, but lack of curation. There is now a desire by internet users for the best information, not any information. There is also a desire for the best entertainment and communication. With some entertainment and information shared, its popularity seems completely random, but is controlled by viral rules similar to natural selection. Media that becomes viral and shared by many people are now called ‘memes’. In this essay, this speaker will investigate meme virality, what types of memes spread and creating local internet content. The web application designed by this student will be used to help investigate these areas. The choices made to create this web application will also be explored.

Meme Virality

This student documented the growth of all three hashtags on my web application, but this student is going to discuss only the #WooliesWaterChallenge in this reflective document. This hashtag started when a group of young men posted a video of themselves drinking Woolworths water, which prompted them to change their accent after taking a sip. Many people created their own versions of the original video, including Mmusi Maimane, the Leader of South Africa’s opposition Democratic Alliance (DA) political party. The original video now has more than 6,500 retweets and 39,000 views on Twitter. How this hashtag become so popular across South Africa will be investigated in terms of meme theory. Limor Shifman, in his book Memes in Digital Culture, says that memes may best be understood as pieces of cultural information that pass along from person to person, but gradually scale into a shared social phenomenon [3]. According to Richard Dawkins, in his book The Selfish Gene, memes that spread successfully incorporate three basic properties—longevity, fecundity, and copy fidelity [1]. All three are boosted by the Internet. Online meme spread has greater copy fidelity (that is, accuracy) than communication through other media, since digitisation allows perfect information transfer. Fecundity (the number of copies made in a time unit) is also greatly increased— the Internet facilitates the rapid distribution of any given message to numerous people. Longevity may also improve, as information can be stored indefinitely in numerous archives on the internet [3]. #WooliesWaterChallenge has all three properties. The hashtag is stored on Twitter and will therefore be stored on servers forever – longevity. Twitter has the retweet function, allowing for a fast spread of viral memes. Twitter also copies tweets perfectly when retweeted, allowing many people to see and react to the same original video. Therefore, Twitter is the perfect environment designed for memes to go viral. Shifman also goes on to note some key attributes when analysing digital culture. These are: (1) a gradual propagation from individuals to society, (2) reproduction via copying and imitation, and (3) diffusion through competition and selection [3]. Again, Twitter is a perfect breeding ground for this viral spreading, as individuals can create content that will be spread over to a wide audience through perfectly copied retweets and replies. South Africa is also a fertile setting as there are many new internet users in the country, but not as much quality content. Therefore, this hashtag was spread without much competition for attention from the South African Twitter audience in March 2019.

Predicting Virality

This student also has some ideas for how this specific hashtag spread and how to predict if a post will be spread and receive a lot of attention before it is even posted. The most obvious being the original account who posted the Tweet. If they have many followers and/or are an official organization, such as a sports team, their tweets have a higher chance of reaching a large audience and being retweeted often. An example would be when Mmusi Maimane (@MmusiMaimane) posted a video relating to the challenge and got over 1500 retweets and
over 1250 replies. The next key ingredient of a viral tweet in the media. This student manually entered all the tweets containing the #WooliesWaterChallenge hashtag and found that if a tweet had an image or video, it was much more likely to be retweeted (see figure 1).
The next factor that affects the likelihood something will become viral is the quality of the tweet and media. If the tweet shows timely comedic insight, effort and engagement, people will want to interact with it and retweet it. An example of this would be the video posted by
@deesorceress a few days after the original #WooliesWaterChallenge video, which received over 245 retweets and over 20 replies. This video showed a spin on the original video after a boy takes a sip of the South African Maheu drink and his accent changes. The final factor that
affects likelihood of virality is the timing of the tweet and hashtag. If the tweet is made without other competing hashtags and when many twitter users are online (before and after the workday), it is more likely to go viral. Again, this can be proved in the @deesorceress
reply video, which was posted at 6:58AM.

Design choices

This student made certain design choices for their web application to best document the spread of the #WooliesWaterChallenge hashtag. The first decision was to limit the number of tweets made for the hashtag to only March 2019 and why there were only about 150 tweets
showcased, but many more about the challenge on Twitter. The tweets were only recorded in March as most tweets were recorded in this time and further growth past this point does not lead to meaningful insight. Secondly, only about 150 tweets were showcased as these tweets were the only ones that were geotagged. Many others had no location information and would be useless to show on the map. It was felt that the average Twitter user would randomly choose to specify their location or not and so the location data would not be skewed because
of this design choice.

Media and retweet correlation graphed using Microsoft Excel by this student

Another design decision this student made was to use the Google Maps Embed API and notthe Google Maps JavaScript API. This was as the Embed API allowed me to create the exact map this student envisioned from their wireframe relatively easily, while a similar solution
could not be found through the research this student did on the Google Maps JavaScript API. Using the Embed API also allowed me to upload Excel files this student manually created of the tweet information. This student did this manual entering as Twitter has severely restricted access in the Twitter API recently. This was done to protect the privacy of Twitter users. Unfortunately for this student, this meant this student could not simply download the desired tweets. This student felt their manual solution was the best one for this obstacle.
An additional aspect of the web application this student wants to discuss is the log graph seen on the right side. This student graphed the cumulative tweets over time along with a similar log graph. This student wanted to show how the growth of the tweets was like the growth of a virus. This similarity can be seen in the actual viral growth graph on the right.

Cumulative growth of tweets (green) along with log graph (red) [top]. Virus growth [bottom] [2]
The final design aspect this student wants to discuss is the explanations of each tweet below the main map. This was to give context to people viewing the website who had never heard about the challenge. This increased the accessibly of the website. The reactive design for different screen sizes also has the same effect.


This project was an interesting way to explore South African viral content on the internet. This was done by tracking and analysing South African hashtags on Twitter using the Google Maps Embed API. Reasons why certain memes become viral on the internet and how to predict if a certain post will become viral were explored. Design decisions for the web application for the South African viral content were also expanded on. This student believes these hashtags are a good step in the right direction of having more South African content on

the internet and the development of the creator mindset local internet communities need to have.


[1] Dawkins, C.R. 2016. The selfish gene Oxford: Oxford University Press.
[2] OpenStax. n.d. Microbiology. Available: nhttps://courses.lumenlearning.com/microbiology/chapter/the-viral-life-cycle/ [9 June 2019].
[3] Shifman, L. 2014. When Memes Go Digital Memes in Digital Culture. DOI: 10.7551/mitpress/9429.003.0005