We frequently see CrowdTangle data being used by journalists all over the world. Here's a quick primer on some of the ways we've seen it used, with examples. We hope this helps give you some ideas, and of course please share if you do a story!*
#1 Surface top influencers, or curate top posts around a particular topic or event
Examples: "Watch the 10 Most Popular Late-Night TV Clips of 2014" by TIME.
How to? Use search leaderboards for the former, and use "search within" filters for the latter.
#2 Analyze what topics are driving the conversation around an event, a cultural moment, an election, etc.
Example: "Midterm messaging revolved around the economy — until the Kavanaugh hearing" by Marketwatch
How to? Use saved search, and then get analyze it using historical data.
#3 Who's "winning" X, Y, Z on Social Media
Examples: "Who's Winning the Social Media Midterms" by the New York Times, "Social Media Ranking: Cruz Azul vs América, who has the best fans?" by Dinero Imagen, "Barca, King of Social Networks" by La Vanguardia.
How to? Use leaderboards, Intelligence, and of course, historical data.
Top to bottom: Axios, New York Times, Dinero Imagen, La Vanguardia graphs:
#4 Coverage of the social platforms themselves, including the spread of misinformation, popular news, communities that are bubbling up, how people are communicating during natural disasters, etc.
Examples: "Inside the Hyperpartisan Fight for your News Feed" by Buzzfeed News, "If a Scientific Conspiracy Theory is Funny, that Doesn't Mean It's a Joke" by The Verge. Screenshots from these stories posted below, in order listed above.
How to? Most of these articles utilized Intelligence and the Link Checker Chrome Extension.
BONUS: Read our Q&A on how Buzzfeed News' Craig Silverman uses CrowdTangle in his day-to-day reporting work.
*Please be sure to consider the context in which CrowdTangle metrics appear. CrowdTangle tracks public posts across 3m Facebook Pages, Public Groups and Verified profiles, 1m public Instagram accounts, and 15K subreddits. We do not include paid ads unless those ads began as organic, non-paid posts that were then “boosted” using Facebook’s advertising tools. We also do not include activity on private accounts, or posts made visible only to specific groups of followers. For more on the metrics we track and how we calculate them, please see our full glossary. And remember! Correlation does not equal causation. A variety of explanations may be available for trends observed in CrowdTangle. Please be sure to caveat your analysis accordingly.
Here's a snapshot:
Overperforming
Overperformance is calculated by benchmarking how many interactions that account’s posts usually get after a certain period of time. Benchmarks are calculated from the last 100 posts of each post type (photo, video, link, etc.) from the account.
Underperforming
Underperforming is a relative score as to how posts for that Page or account normally perform. If it’s red, it's performing below average.
Total Interactions
The total number of likes, reactions, shares and comments on a Facebook post; the total number of retweets and likes on a tweet; the total number of likes and comments on an Instagram post; the total number of upvotes and comments on a reddit thread.
Interaction Rate
Calculated by averaging the number of interactions for all of the account’s posts in the specified time frame, then dividing that by the number of followers/fans.
Weights
The customizable weights given in our algorithm to each interaction (for example, a reaction, share or comment). Choose which interaction is most important to you.
Video View
The number of times a video was watched for an aggregate of at least 3 seconds, or for nearly its total length, whichever happened first (more info here). However, CrowdTangle updates video views captured every 4-6 minutes, so the numbers reflected in CrowdTangle and what you see on Facebook may occasionally be briefly out of sync.
Have questions on a story or need more data? Email press@fb.com.