I was convinced that the referendum decision on 24 June would be for Britain to remain in the European Union.
It seemed that almost everything I read was in favour of the status quo.
In fact social media called the result of the EU referendum correctly. Like many people I wasn’t looking hard enough.
Blowing Facebook filter bubbles
The Facebook newsfeed reinforces existing biases. It isn’t a good place to start if you’re seeking to be informed about political discourse.
The media that you’re served in the newsfeed is based on how you interact with articles posted in the newsfeed.
If you liked and shared content supporting Britain remaining in Europe during the lead-up to the election, that’s what Facebook’s algorithm would serve you whenever you visited the social network.
It’s called a filter bubble as it lulls you into a false sense that the world is aligned with your point of view.
I should know better. I’ve written about this issue before.
Critics call Facebook an affront to democracy but media has always operated this way. We’ve traditionally made media choices on the basis of our comfort with a particular bias.
The size and scale of Facebook is a concern. According to Facebook Power Editor there are 37 million voting adults in the UK using the platform.
However, if you’d delved deeper into Facebook you would have found greater sources of insight into the political debate.
Communities as a source of influence and insight
I’ve long argued that online communities are one of the most influential forms of media. People come together to discuss topics and issues.
Communities exist on almost every form of social media ranging from Twitter to Wikipedia, and from Reddit to Instagram.
Within Facebook, communities operate within groups. These are places that community members visit to read content that has been posted and participate in discussions.
Groups are a place to visit and be informed rather than engage in discussion.
Views are typically polarised and behind the mask of a screen and keyboard people are more direct and often ruder that they would be in real life.
In the lead up to the referendum, Hotwire’s director of engagement John Brown spent 24 hours exploring groups for and against Britain’s exit from the European Union.
Brown was left with a clear view that Britain would vote leave.
“Brexit groups were ardent and passionate. The economy and immigration were frequently discussed issues,” said Brown.
“But there was something more fundamental at play. The referendum was an opportunity for the country to voice discontent at the political classes in London and Brussels,” he added.
Brown’s view is consistent with analysis from Brandwatch. Its analytics tools analyse data from blogs, forums, Twitter and public areas of Facebook.
I also experienced a filter bubble on Twitter. My network was firmly in favour of Britain remaining in the EU.
Two weeks ahead of the referendum, Brandwatch reported that the use of vote leave hashtags was ahead of vote remain by almost two to one.
There’s no correlation between hashtag usage and voting intention however, like many people, I was blind to the strength of the opposition debate.
What have I learnt? Modern public affairs needs to start within the communities that campaigns are seeking to engage.
Hotwire’s Brown has started to attend political meetings. He believes this is the only way to truly engage in informed debate.
If you’re looking to Facebook for answers you need to explore both sides of a debate in groups. Similarly, on Twitter explore opposing hashtags.
Pop filter bubbles created by algorithms and networks. It’s critical to understanding democracy.
Stephen Waddington is chief engagement officer at Ketchum and visiting professor in practice at Newcastle University. He tweets @wadds