#happy is… is an exploration of happiness on social media. Through this project, I investigate what happiness looks like according to our social media habits, who is sharing photos tagged as #happy, and from where they are being shared.

ABOUT THE PROJECT

Since 2013, I have searched for the best ways to collect information about happiness on social media. The  #happy is… project focuses solely on Instagram because it is first and foremost a visual platform that allows users to share photos and videos. Over 289 million #happy photos are available on Instagram, out of 40 billion total Instagram posts, with thousands and thousands of #happy photos added every day.

Over the course of 12 months, I collected a data sample of over 100,000 public photos tagged as #happy. Each Instagram post contains much more information than the image — information called metadata that makes up more than 20 individual fields such as username, ID, tags, filters, likes, longitude, and latitude. My data set quickly jumped from 100,000 photos to over 2 million individual data points from which I could extrapolate and analyze trends.

For the past two years, I have created artworks from the data I collected, and exhibited works as fine art prints throughout the country. In the winter of 2015, I published a book that contains prints, data statistics, and the process behind the #happy is… project.

INSTAGRAM

1
Monthly actives
40
Photos Shared
3.5
Likes Daily
80
Average Photos/Day

Process
01. Data

IFTTT and Zapier, both free online services that automate simple conditional statements, are used to collect data. The statement used for #happy is… is quite simple: If anyone creates a new public Instagram photo tagged as #happy, then upload the Instagram file to Google Drive and create a new row in a Google Spreadsheet to document metadata associated with the photo.

Process
02. Analysis

The information from the data set is then parsed into smaller subsets such as image, caption text, tags, filters, latitude, longitude, likes, and comments. Initial data analysis is completed through Google Drive and Microsoft’s spreadsheet tools to discover trends, such as which filters are most popular for #happy photos, how many likes #happy photos average, and where #happy photos originate.

Process
03. CODING

Processing, an open-source computer programming language, is used to further investigate and visualize the data. Processing is “a flexible software sketchbook and a language for learning how to code within the context of the visual arts.” The artworks and prints within the #happy is… project were all coded and created with Processing.

14
Average likes per #happy photo

#happy posts applied with a ``Normal`` filter

66%

Instagram users who live outside of the U.S.

75%

Initially, the #happy is… project started by looking at the most common color, and sometimes the frequency of that color, of an Instagram photo. Over time, I began analyzing all of the other metadata associated with an Instagram post and found that the average #happy photo receives 14 likes and that 66% of #happy photos are published without an Instagram filter.

Interested in more data behind the #happy is… project? Please consider ordering a limited first-run print edition!

Special thanks to my wife Heather, for our incredibly happy lives together, and to all of those close to us who asked if we were happy after our engagement… it led to great things!

Thank you to my family and friends for your support, especially Kevin and Diane Cavanaugh, Jen and Jeff Baney, Ismael Cervantes, and the Kickstarter backers who made this project possible.

This project is sponsored, in part, by the Greater New York Arts Development Fund of the New York City Department of Cultural Affairs, administered by Brooklyn Arts Council (BAC).