Another General Election is over. Joy for some, despair for others.

For many, total and utter shock at the result. What happened to the grassroots Labour support? What happened to #YouthQuake?

At team Fluence, we were naturally quite curious about the disconnect between what people saw on social media and the cold reality of a ballot box.

Seeing that we are having our Xmas party this afternoon, we figured that there was no better way to spend our morning than by dissecting the Twittersphere?

Before giving you any of our findings, I should point out that we performed this analysis in a very short timeframe, which means that our data sample is relatively small (around 10,000 Tweets). Having said this, our findings were quite interesting. Also, as we only had a couple of hours, and we can’t speak Welsh, we omitted Plaid Cymru from our analysis (really sorry Wales).

Labour Vs Conservative intent:

As it turns out, based on Twitter opinion alone, Labour won by a landslide. The ratio of political intent is massively skewed toward Labour, with 4 in 5 ‘explicitly pro-party’ Tweets landing on Labour’s side. With this in mind, it is not so surprising that the Twitterati were convinced of a different outcome.

 

Understanding party voices:

Firstly, we wanted to understand the relationship between the formal Twitter campaigns, which we did by analysing official party Twitter accounts (i.e. @theSNP, @brexitparty_UK, @Conservatives, @Labour, etc.) as well as the accounts of the most prominent figures within each party (i.e. Boris Johnson, Jeremy Corbyn, etc.)

We analysed the relative position of each party in the Twittersphere. And as it turns out, Labour had a much stronger Social Media showing, with their campaign sitting in a more central position on the map and capturing a larger share of the language we analysed. However, Labour’s formal Twitter channels were sat quite closely between Lib Dem messaging and Green Party messaging. It also suggests that Labour’s Twitter activities were significantly more diverse than their primary opposition, the Conservatives. Our working hypothesis is that this may have diluted the conversion rate of Labour’s messaging. This essentially means that a Labour Tweet was also promoting the ‘campaign values’ of the Green Party and the Lib Dems.

However, unlike Labour, it seems that the Conservatives, SNP and even the Brexit party enjoyed a greater level of differentiation from their opposition. This suggests that these parties ‘owned their space’ in a more convincing way. This means that any message released by one of these parties had a greater chance of leaving a ‘party-specific’ impression on the audience.

 

Conclusion:

It is probably best not to rely on Twitter for political validation. It seems to present as a classic example of selection bias, and it does not seem to reflect the cold hard reality of public opinion.