2016 | Google News Labs

WorldPOTUS

Interactive, Dataviz, Storytelling, and Experience

Who would the rest of the world vote for? A data-driven digital experience visualizing “soft” and “human” data to discover the world’s interest in the 2016 US presidential candidates.

Brief

In the Summer of 2016, Simon Rogers, Google News Labs’ data editor, and Alberto Cairo, the Knight Chair in Visual Journalism at the University of Miami, contacted Accurat to lead one of the projects in a series of experiments to work with renowned global data visualization teams to find new ways of using Google data in newsrooms. The 2016 US presidential election was approaching and interest around it was spiking. The president of the United States is often described as the “leader of the free world” and this de facto leadership is indeed perceived directly or indirectly in many corners of the globe, but very few ask or measure how the rest of the free world perceives this. Starting from this very simple concept, Accurat proposed to use Google Trends data to measure the interest in the candidates in relation to top political issues, but outside the US.

Accurat conceptualized, designed, and developed both the back-end and the front-end of WorldPOTUS, a React and WebGL interactive web and mobile application that displays how people around the world search for the key issues of the election associated with the two candidates, Hillary Clinton and Donald J.Trump.

Solution

'By looking at searches for each candidate and how they differ in each country, WorldPOTUS shows real-time worldwide variations in searches — and which region is interested in each of the major election issues. The topics — Abortion, Black Lives Matter, Brexit, Economy, Foreign Policy, Global Warming, Gun Control, Healthcare, ISIS, Oil, Racism, Refugees, Same Sex Marriage, TPP — are selected based on the top issues of the campaign as decided by their ranking in Google search on average in 2016. But, why these wobbly bubbles? Shouldn’t data visualization always be rigorous and precise? The Google Trends Index is a valuable and insightful way to gauge the general sense of a country’s interest in a topic over time, but it shouldn't be considered a perfect and ultimate measure of it. Opinions on such delicate topics are often a fluid and ever-changing blob and, for this reason, Accurat worked on a “softer” depiction of data, in a way, that faithfully renders the fluidity and fuzziness of conversations around the world – a deliberately imperfect visual to grasp and display a meaningful and rich, but very human, dataset.'

Accurat imagined an original data visualization that transforms countries into gooey bubbles, bouncing around as the user switches from topic to topic and sticking to each other depending on additional information on search habits. Accurat’s goal was to represent this “soft nature” of data on people’s interest in the topics of the US election in a way that could convey the general trends and magnitude of the phenomenon, rather than the precise number represented by the index.

The project is based on React framework and WebGL. Code is opensource and available on GitHub.

Process

During the development of the app, the team faced a number of fascinating technical challenges. The primary one was smoothly rendering on multiple devices the gooey blobs that represent the search interest relative to all world countries. Most of the existing tests and examples were built using SVG filters but, through experimentation, Accurat found that WebGL shaders have wider browser support and provide much better performance for these type of effects, while also providing a nicer and smoother result in the end. On the data side, another interesting challenge was calculating meaningful continental aggregates, based on the Google Trends index, available only at the country level: how do you faithfully represent the search interest for a continent if only a few countries of that region have a relevant value in the index? To do so, we weighted the Google Trends indexes of each country by population and percentage of active internet users, and we used the results to determine a value that would accurately represent the weighted average interest for each world region.
The app is built around three independent but complementary views that differ based on how the bubbles are arranged in space: although each one is able to be digested individually, they provide the best insight while seen together. BLOB, ATLAS, and DROPS, are all based on the same primitive blob-like shape that melts and evolves into more complex, organic structures during navigation.
BLOB - the proximity of the bubbles is based on similarity in search habits across all topics.
ATLAS - countries and world regions are organized by geography, divided into five areas: Americas, Africa, Asia, Europe and Oceania
DROPS - bubbles are distributed vertically, depending on the past twenty-four-hour trends in search interest.

Results

As a result of Accurat’s work, WorldPOTUS was featured prominently on the homepage for the Google Trends story on the elections as one of the main tools journalists used to report on the global perceptions regarding this ever-evolving and often controversial topic. The project was covered internationally by major news outlets such as Wired↗ and Fast Company↗.
Team
Giorgia Lupi
Gabriele Rossi
Simone Quadri
Amin Al Hazwani
Marco Bernardi
Davide Ciuffi
Paolo Corti
Luca Falasco
Pietro Guinea Montalvo
Tommaso Renzini
Cesare Soldini
Marco Vettorello
Tommaso Zennaro
Francesco Costa
Simone Tolomelli
Alberto Cairo
Simon Rogers
Jennifer Lee
Press
Wired ↗
Here’s What the Rest of the World Is Googling About the Election.
Fast Company ↗
Google Searches Reveal How The Rest Of The World Sees This Terrible Election.
Services
Interactive
Dataviz
Storytelling
Experience

Interactive, Dataviz, Experience, Big Data, and Visual Analytics

Reporting, Dataviz, Editorial Design, and Interactive