The Post Heard ‘Round The World

    You and I are similar… even if we don’t live in the same country. An interactive visualization on global traffic.

    Thanks to the incredible power of social networks, the world feels a whole lot smaller than it ever has before. Here at BuzzFeed, we create and translate posts daily. These then propagate to users around the globe. But what pattern does this content-sharing follow? And what makes a post go viral internationally?

    To start answering questions like these, we built a tool to visualize the spread across geographical boundaries. Along the way, we discovered that for popular posts (those with hundreds of thousands of pageviews), like 31 Sentiments that Every Dog Owner is Familiar With, city-dwellers' shared interests spanned miles and oceans.

    Emerging Patterns

    "31 Sentiments" is first published in New York City in early February, then rapidly spreads to the UK, Singapore and Australia in just two days. Almost immediately afterwards, the French translation is published, yet just a few, small orange bubbles appear in cities outside of Paris… suggesting that perhaps most users in French-speaking countries are uninterested -- or aren't dog-owners. And by mid- to late-February, Portuguese- and Espanol-translated posts explode in popularity, appearing in South America in huge numbers.

    This post is just one of several pieces from 2014 that see a huge peak in pageviews immediately after publishing. But unlike other posts, "31 Sentiments" saw another, smaller peak when other translations were published. In fact, this "second wind" appears in just some of the most popular posts, suggesting that only super relatable topics lead to viral paths.

    The conclusion? Shared, common experiences lead to urbanites feeling "closer" to other city-dwellers in different countries.

    A Note about the Data

    For this post, we studied over four-hundred thousand (anonymized) pageviews at a city level. The data, pulled and merged from Google Analytics and our internal SQL store, was cleaned and transformed in Python. The interactive below was built using D3.js.

    Each bubble's area is sized proportionally to the total number of pageviews, with a threshold of at least forty total pageviews per city. Overlapping colors indicate dominant translations of the post, and are in one of three other languages: Spanish, French, or Portuguese. Consider Madrid, for example, where both the Espanol article and the original English version are popular, resulting in blue and pink bubbles.

    The panels chosen for display are countries with the highest BuzzFeed traffic, which precludes continents like Africa and Antarctica, and countries in the Middle East and eastern Asia. Color changes occur on the date that foreign translations are published, or when cities begin rapidly consuming that content.

    Check out the interaction below (and mouse over your city to see how it performed)!

    31 Special Feelings That Every Dog Owner is Familiar With

    Anita Mehrotra / Via BuzzFeed Data github.com

    Code for the visualization can be found here. A special thank you to Taylor Baldwin and Jeremy Singer-Vine.