The influence of modern technology and engagement in cyber socialising have become a prominent part of modern communication. A new learning pedagogy with proper guidelines is needed to assist users to engage with social networking platforms efficiently. The researcher investigated discourse analysis involving participants to answer questions about both the contextual application of the language, and the functions and results of aspects pertaining to discourse such as diction, cohesion, and metaphors. The researcher employed a quantitative approach in this study. The study randomly sampled 80 students from a University of Technology (UoT) in Gauteng to participate in quantitative discourse analysis of communication using Facebook, WhatsApp and Twitter texts. The methods of data elicitation embraced extracts from Facebook, WhatsApp and Twitter texts provided by 80 participants. Statistics were used to present the findings of the quantitative data which included mainly frequency of certain aspects pertaining to discourse. Data were collected via email and texts were numbered according to the participants. These texts remain anonymous and the identities of the participants were concealed. The study found that the language used on Facebook, WhatsApp and Twitter was characterised by the use of emoji, low register words, code-switching few spelling errors and the modern tendency to shorten words by using clipping and number homophones which are not seen as errors but a unique style of writing. Facebook, WhatsApp and Twitter remain three popular communication sites and discourse surfaces since it is important for users to communicate meaningfully even when using excessive punctuation to indicate excitement and emoji to communicate emotions.