Social media recorded drop in Kenyatta support
BrandsEyes’ latest research between August 28 and September 4, looked at online conversation from 150 000 Kenyans in order to gauge public sentiment towards their President Uhuru Kenyatta and opposition leader and presidential candidate, Raila Odinga. The findings show Kenyatta’s attack on the Kenyan judiciary angered Kenyans and led to a further drop in support for the incumbent leader.
In response to the Supreme Court’s 1 September decision to annul the election results, Kenyatta referred to the judges as “wakora” – meaning crooks or thugs.
Kenyatta experienced a peak in net sentiment on the August 31 as the Supreme Court investigation into the recent Kenya presidential elections concluded. Kenyatta’s supporters rallied around the candidate calling on the Supreme Court to “respect the will of Kenyans”.
The Supreme Court’s decision was positively received by Kenyans who viewed the announcement to suspend the result as a triumph for democracy and rule of law.
The Supreme Court’s decision was positively received by Kenyans who viewed the announcement to suspend the result as a triumph for democracy and rule of law.
In the week leading up to the Kenyan elections (1-6 August 2017), 443 000 mentions from 173 000 Kenyans discussing the election were analysed. According to the data, BrandsEye predicted that incumbent Kenyatta was under threat of losing the presidency.
Prior to the court’s intervention, BrandsEye’s data shows that there were 12 937 mentions pertaining to vote rigging and fraud on August 9, the day after the election. This points to some concern from voters about the fairness of the election.
Of those 12 937 mentions, 61% referenced Odinga’s claim that the election system was hacked.
Prior to the election results being announced on August 11, Kenyan social media signaled that fraud had played a role in the voting process, highlighting the predictive value of social media as a gauge of public opinion.
BrandsEye method
Social media opinions have traditionally been very difficult to analyse in aggregate due to the scale and complexity of the conversations. The scale challenge can be solved using artificial intelligence (AI), but this approach leads to inaccurate data, which at best is 60% as machines struggle to understand the nuances of human conversation.
BrandsEye solves the inaccuracy challenge by complementing the AI with trained crowds – large teams of local language speakers who review and verify the data allowing us to get online accuracy of 97%. In the case of this election, our crowd was made up of Swahili speakers who understood the local dialects and social context.
Kenya’s presidential elections have been rescheduled for October 17.
