Back in June, I predicted that the Leave camp would narrowly win the Brexit referendum through digital data. The analysis looked at 3 data sources including demand, web traffic and social mentions, and showed a small lead for Leave both in terms of traffic and social media volumes.This begs the question: can we truly predict election results through digital data?
The US Presidential elections offer the perfect opportunity to test this methodology again, scrutinizing a highly digitally literate market. In the US, 78% of Internet users in the USA had a social media profile in 2016, a 5% growth versus the previous year, reinforcing the importance of social media as a communication tool. Additionally, Internet penetration in the USA is estimated at 88.5% suggesting that the majority of US voters may have engaged with the Internet to inform their voting decision.
How are Trump and Clinton performing digitally in 2016?
Similarly to my previous analysis of Brexit, I looked at 3 digital sources: demand, traffic data and social data.
Findings show that Clinton has received a higher traffic share than Trump overall, and despite having a smaller social footprint, interestingly received a considerably higher share of social traffic to her site. It would appear that voters are cognitively engaged with her content, leading to clicks. Then again, the amount of social traffic would also correlate with the quantity of links shared throughout the candidates’ respective campaigns.
Although Trump wins in terms of volume of mentions, sentiment around the controversial candidate remains largely negative. Saying that, Clinton also received high levels of negative mentions overall. In terms of consumer emotions Clinton inspired ‘disgust’, whilst Trump was found to predominantly inspire ‘fear’… The private email servers saga certainly didn’t help Clinton’s campaign as seen through Google demand search terms and voters’ emotions. Overall, however, Trump received higher volumes of mentions from voters pledging their support (e.g. Vote Trump), although this could be down to Twitter bots.
In sum, Clinton was found to beat Trump in terms traffic data, website engagement and but also social media sentiment, whilst Trump had higher levels of voters pledging their support.
Thus, based on these data, it may be inferred that Clinton is set to win the Presidential election.
Take a look at the presentation below to view my findings.