Could social media be used to call Europe elections?

This article was first published on Newsweek.

Written by JP Kloppers, DataEQ (former BrandsEye)

Traditional polling failed to predict Brexit or the election of President Donald Trump, but an accurate understanding of sentiment on social media saw them both coming. The reason is simple, but often difficult for political elites to deal with: We have entered a new age of identity politics and backlash to globalization.

These factors are better understood by mapping out the emotions found in discussions on social media; we cannot rely any longer on sterile dialogue between sample groups and pollsters.

Win an iPhone 7 Sign up to our daily newsletter for your chance to win. This year, both France and Germany will elect new leaders. In both countries, right-wing populism is on the march, fueled by concerns surrounding immigration, terrorism and the EU. The similarities between this politics of patriotism and nationhood and the success of Trump in America are clear. However, it is unclear whether we have begun to learn any of the lessons of 2016.

First, it is remarkable to watch the speed with which we have returned to polls as a guide and political forecast. Every major media outlet and pollster was caught out by the Brexit result and Trump’s election to office. From the BBC to CNN, Gallup to Nate Silver, the polls missed the point of these landmark votes.

There is a lot of talk about the “post-truth” era we live in, but I believe a better phrase is the fact that we live in a “post-trust” era. Trust is an emotional judgment, whereas truth is an argument of fact. Polls fail to evaluate the emotional context of voting decisions. What people feel today affects what they do tomorrow, and traditional polling simply cannot capture this.

While traditional polling is by and large reliable, it does suffer from two structural problems: firstly, it cannot measure the strength of emotion, and secondly, it can’t measure unsolicited and spontaneous opinions. We have also seen that people can be unwilling to admit sympathy for the Tories, or Trump, or Brexit, to a pollster over the phone—even if they intend on voting according to that sympathy.

And yet, pick up any paper this week covering the French election and you will read about Marine Le Pen’s strong poll figures and the surprise surge of outsider Emmanuel Macron in the latest polls. There seems to be a concerted effort to keep up the echo chamber that left us scratching our heads in 2016.

This brings us to our second lesson. If we are to truly gauge popular sentiment and understand how our institutions are viewed and why, we must burst the political and informational bubble that we have grown so comfortable in.

Social media offers a route out of the echo chamber. The millions of online conversations taking place every minute offer both the quality and quantity of data to get to grips what people are actually thinking. The challenge, however, in harnessing the insight that intrinsically exists in this dataset is that even the best artificial intelligence in the world can’t accurately decipher the meaning and emotion expressed in unstructured text.

Social media laid out the raw emotions that fueled the election of Trump and the UK’s decision to leave the European Union. At DataEQ, the media analytics agency where I am CEO, we focused on uncovering and understanding these emotions using our crowd-sourcing technology. This meant we were hardly surprised when voters made decisions based on nostalgia and national ties, as opposed to stability and economics.

There is plenty that leaders can do to avoid reproducing last year’s political surprises and the disconnect that has emerged between elites and electorates. The shocks of 2016 were not at all shocking if political leaders had listened to the voices of the people, and thus understood this new strain of populism within Western politics.

Emotion has always been a political currency. An ability to shape it, through compelling campaigns and inspiring narratives, has driven the victories of our generation, from Tony Blair to the Brexiters, Trump to Barack Obama. The upsets of the past year have shown that telling stories, rather than stating facts, will win you elections.

Hillary Clinton lost because she did not speak using an overarching narrative—her policy advisers pushed out position papers while Trump sold trucker hats promising to “Make America Great Again.” While the “Leave” campaign spoke about patriotism and nationhood, the “Remain” camp discussed trade deals and GDP.

In this age of renewed questions around national identity, globalization and liberal democracy, politicians must understand hopes and fears, not checks and balances.

2016 is full of warning signs for both Merkel and Fillon, as well as for the establishments behind both of them, who simply cannot countenance a populist and nationalist France or Germany. But all is not lost. Understanding what people are thinking and why is as simple as engaging with the conversations that are currently playing out on social media.

Using this snapshot of popular sentiment can inform powerful campaigns. We all remember “Yes We Can” and “Education, Education, Education.” Both the French and the German elections will be won by storytellers. The hope for the liberal consensus is that it is the right story being told.

Jean Pierre Kloppers is CEO of social media-driven analytics agency DataEQ.