The Council for Scientific and Industrial Research is working closely with SABC News during the Elections period, not only calculating predictions for the race for votes and race for seats, but also providing a wealth of data relating to the social media side of the elections, specifically on Twitter.

For example, what are the most-mentioned elections-related handles? Who are the most popular politicians and parties on Twitter? From which areas are people tweeting most about elections?

The following table shows the total mentions of the top three parties (not just their official handles) and their respective leaders for the period from the morning of Monday 6 May until 9pm on Thursday 9 May. The EFF comes out on top with 18 969 mentions, the ANC received 17 132 mentions and the DA received 11 843 mentions.


The following chart shows the number of mentions of the top three parties’ official handles, namely @MyANC (in green), @EFFSouthAfrica (in red) and @Our_DA (in blue), for the 12-hour period from 9am on 9 May to 9pm on 9 May.

For most of this period, the EFF’s handle was tagged more frequently than the ANC’s and the DA’s, but from 5pm the conversation shifted and the parties were more closely matched.



The next chart shows the number of mentions of each of the leading parties’ top candidates: @CyrilRamaphosa of the ANC (in green), @Julius_S_Malema of the EFF (in red) and @MmusiMaimane of the DA (in blue) for the same 12-hour period from 9am on 9 May to 9pm on 9 May. Malema received the lion’s share of mentions for this period, peaking at over 60 in a 10-minute period at around 7pm.


For the same 12-hour period, these were the most popular elections-related handles mentioned on Twitter. @EFFSouthAfrica leads the way with around 4 000 mentions, followed by @IECSouthAfrica with around 3 100, and @Julius_S_Malema with around 2 600.

For the same 12-hour period, the following areas in South Africa saw the most number of tweets containing the #SAElections2019 hashtag. Note that this data is restricted to tweets that were sent from devices that had their GPS location enabled for Twitter, so may not be an accurate overall reflection of where people were tweeting from:

For the same 12-hour period, the following were the most frequently used hashtags related to the ANC, DA and EFF respectively:

Note that all of the above figures (and their associated charts) rely on data acquired via the Twitter API, which sometimes limits data requests according to server load. The figures may therefore not always be 100% accurate, but do offer a very good indication of the proportionate relationships between keywords and handles over time.