No correlation between social shares and people reading anything

Paul Dughi
5 min readJun 20, 2016

When we post a story about something that happened in our town on Facebook, we almost immediately see people sharing it and commenting on it. More often than not, somebody asks “Where did this happen?” or “Is everybody OK?” In almost every case, that information is contained in the article. They’d know that if they just click on it.

We’ve all seen posts that quickly move off the actual story and evolve (or devolve as the case may be) into a debate about the issue (and not the facts of the story) or sometimes veer off into weirdness. Every time we post a story that has anything to do with politics, or terrorism, or — god forbid — gun control, it’s clear most of the commenters have never read the story.

How common is it? A new study by computer scientists at Columbia University says that as many as 59% of the links shared on social media have never been clicked. Most people are retweeting and/or commenting on something without ever reading the underlying item.

So it’s not really about what the article says. It’s about what the headline infers that generates that emotional connection. Hmm.

You can read the full study here, but it’s full of charts and graphs like this one. Or you could just keep reading this article and I’ll summarize it nicely for you.

The study team tracked five news sites (The New York Times, BBC, Fox News, CNN, and The Huffington Post) on Twitter. In total, they looked at 2.8 million shares which led to 75 billion potential views, and 9.6 million clicks to 59,088 unique resources.

After they tracked where the shares went and how many times a story was actually clicked on, they determined that stories that go viral can be widely shared, but rarely read.

Maybe that “sharing without reading” is how a certain Presidential candidate can tweet THANK YOU!! attached to a story about a poll that showed him losing to his opponent.


Prevailing wisdom, according to the study, is that since a fraction of users clicking an article will decide to share it, the share number is a good way to estimate actual readership and popularity. But the study warns about using shares or retweets as a way to gauge success.

Majority never read what they are sharing

59% of the shared URLs are never clicked (or silent as the study calls them).

Sharing content and actually reading it are poorly correlated

When it comes to popular content, sharing activity is “not trivially correlated with clicks.” The most shared content did not equate with the most clicks. In fact, the viral stories had a smaller percentage of clicks compared to the number of shares.

So that explains why people might see a post about Steph Curry’s wife Ayesha posting a “heat of the moment” tweet saying the “NBA is rigged” during the NBA Finals and a commenter might react saying that her husband was the one that rigged the game.

“People are more willing to share an article than read it…This is typical of modern information consumption. People form an opinion based on a summary, or a summary of summaries, without making the effort to go deeper.” — co-author Arnaud Legout in a statement to The Washington Post

It certainly explains how the Science Post ran a story titled “70% of Facebook users only read the headline of science stories before commenting” and then filled the article up with latin filler words.

That didn’t stop hundreds of people from commenting on the article on social media expressing rants about other people commenting on articles… without reading them first.

In fact, the article was shared 45,000+ times with tons of comments on social media.

Or why NPR can post a story titled “Why doesn’t America read anymore?” as an April Fool’s prank and have tons of commenters spouting off about how awful that is without clicking on the story… which asks people NOT to comment to see what happens.

No correlation between social shares and people reading anything

Chartbeat is in the business of analyzing social shares and web traffic. They look at billions of instances every month across more than 4,000 top publishers. Tony Haile, its CEO, says “We’ve found effectively no correlation between social shares and people actually reading anything.”

Even when they do click, they typically read very little

Chartbeat looked at deep user behavior across 2 billion visits across the web over the course of a month and found that most people who click don’t read. In fact, a stunning 55% spent fewer than 15 seconds actively on a page.- Tony Haile, CEO of Chartbeat via

Haile points to the widespread assumption that the more content is liked or shared, the more engaging it must be. Therefore, people are more willing to devote their attention to that article. “However,” Haile said, “ the data doesn’t back that up.”

SOURCE: Tony Haile, Chartbeat

In fact, it’s more likely the opposite. The less time people spend on an article, they more likely they are to share it. Maybe that’s why “snackables” and “listicles” get so much social buzz.