I am shocked--shocked!--to find out that clickbait headlines are being used on the Internet.

There's an academic study. A trio of researchers from the University of Mississippi and the University of Oklahoma say they analyzed 1.67 million Facebook posts that had been posted by a total of 153 media organizations.

As described in their article, "Diving Deep into Clickbaits: Who Uses Them to What Extents in Which Topics with What Effects?" (.pdf link) they developed a "clickbait detection model," and tried to figure out whether "mainstream media" or "unreliable media" used clickbait more often.

Here's how they describe the method:

Specifically, we use distributed subword embedding technique to transform
the words in the corpus to 300 dimensional embeddings. These embeddings are used to map sentences to a vector space over which a softmax function is applied as a classifier. Our best performing model achieves 98.3% accuracy on a labeled dataset. 

I'm just going to be honest: that's total gibberish to me. I have no idea what they're saying. But that's okay, because the results they came up with are entirely predicable.

In short, they say that both "mainstream media" and "unreliable media" often use clickbait--and that it grew in prevalence between 2014 and 2016. Specifically, 19.46 percent of headlines were "clickbait" under their definition in 2014; 23.73 percent in 2015; and 25.27 percent in 2016.

As one of the study's coauthors, Naeemul Hassan, an assistant professor of computer science at the University of Mississippi, told me in an email, they define as clickbait as "a form of web content that employs writing formulas and linguistic techniques in headlines to trick readers into clicking links, but does not deliver on promises."

And while this whole study makes me laugh a bit--because yes, I think the findings are kind of obvious--it's also important to focus on the two parts of that definition: both (a) prompting people to click, and (b) also not delivering on promises.

Because I think a lot of things get caught up in the general definition of clickbait that aren't really clickbait. They're "clicky," maybe, but not deceptive.

And that in turn, I think, is because most serious digital writers want Internet traffic, but they also eschew dishonesty. They don't want their headlines to overpromise things they know they can't deliver.

(Quick timeout: I'm giving myself a pass on the headline for this article, which was of course intended to be ironic--hope you're finding this useful.)

Here's an example, because I live and breathe these kinds of headline decisions every day. Recently, I interviewed the founders of a litigation finance startup about their strategy of providing funding to litigants in small dollar disputes with big companies.

It sounds kind of boring when I summarize it like that--and frankly, I don't know who would be enticed to read the article if I'd used a sort of New York Times-y headline like: 

In Small Claims Court, Plaintiffs Find a Way to Even the Odds

So, I used something that's a little more clicky: 

Is that unfair? I hope not. It sparked a discussion about the nature of clickbait on Reddit, in fact. Fortunately for me, most commenters defended it.

One of them wrote, "There's nothing wrong with clickbait when the title is accurate." (He or she got upvoted too, so that's a good feeling.)

Honestly, it all comes down to economics--both long term and short term.

Long term, media that don't deliver, ultimately won't be trusted. (Although unfortunately, they often drag others down with them, too.)

But short term, there's nothing sadder for a writer than pouring heart and soul into an article online and finding that nobody reads it. Besides, we're still in a position in which most media are advertiser supported.

I give them credit. It got me to click.

Published on: Mar 22, 2018
Like this column? Sign up to subscribe to email alerts and you'll never miss a post.