Last year, the world of journalism lost long-time New York Post staffer Vincent Musetto, who came up with what is often cited as the greatest newspaper headline in history: “Headless Body In Topless Bar.”
Most reporters don’t want to hear this, but the last thing a reader is inclined to do is invest his or her valuable time reading a story. That’s where the power of the headline comes into play in print. Not only will a great headline help sell a newspaper, it is often the main reason a skimming reader will stop and pay attention to a particular story.
Writing digital headlines is an entirely different game. Headlines can still make or break a story, but online, all a reader has is a headline (and sometimes a photo or piece of art) to make a conscious decision to click into a story.
Thus, the term “clickbait” was created, and suddenly bombastic, sensationalized headlines from outlets like Upworthy and BuzzFeed were able to mine the curiosity gap of its readers to the tune of millions and millions of page views. It didn’t take long before more traditional media companies adorned their websites with those new headline-writing formulas in an attempt to bolster click-through rates.
Now it turns out those clickbait headlines techniques might not work so well after all.
Chartbeat, a popular analytics firm that provides real-time data to hundreds of media companies, has published a surprising study about headlines using data from its new Engaged Headlines Testing. After crunching the numbers, the biggest takeaway is most of the headline writing tricks we’ve been taught and use daily don’t really work.
“We’ve proven that ‘5 ways to write the best headline ever’ isn’t actually that effective,” said Chris Breaux, the data scientist at Chartbeat who put together the report. Using the data from 10,000 headline tests from more than 100 domains, Breaux found there was no significant advantage for headlines that used one of the many fill-in-the-blank gimmicks over a more straight-forward approach.
“The results echo stuff we’re reading form Upworthy and others that maybe the curiosity gap is closing,” Breaux said. “If you have a hook or a technique that works, readers only have so much interest, and overplaying it may not be effective over time.”
The findings also surprised Andy Bechtel, an associate professor at University of North Carolina, Chapel Hill, who has taught classes on digital headline writing. Bechtel was particularly surprised that using a name of a person wasn’t especially effective in driving more clicks.
“I’d still recommend doing that, not only for SEO purposes but also because ‘Area Band to Play Local Festival’ is vague,” Bechtel said. “Be specific.”
Chartbeat isn’t alone in proclaiming an end to overused words and paint-by-numbers gimmicks. An analysis of more than 3.3 million headlines from more than 100,000 websites by Hubspot and Outbrain showed headlines with words like “you,” “easy” and “how to” also underperformed compared to a more direct headline.
Instead of searching for a silver bullet, Breaux’s study suggests the most effective way to come up with a successful headline is to simply A-B test as many alternatives as possible and let readers tell you which headline works best.
According to Chartbeat’s data, in a single A-B test between two headlines, the winner generated about 23 percent more clicks. But editors who were able to test six different headlines generates 2.6 times more clicks than the average of those tested, a rift that increases the more headlines you’re able to test.
Other best practices for A-B testing include testing content in higher traffic positions, as smaller numbers can tend to skew results. Also, editors should test headlines that are distinctly different from one another.
“One common mistake is producers will take a headline and switch two words around,” Breaux said. “As you’d expect, it won’t lead to a large shift in clicks there.”
Bechtel agrees, noting the data gleamed from multiple headline testing is valuable not only in generating clicks, but devising a strategy for content moving forward.
“The beauty of digital headlines is we can gather data on what works and what doesn’t, and we can update and revise as much as we want,” Bechtel said. “In print, we never knew what readers liked or disliked, and once a headline was in print, that was it.”
Despite the clear shift away from headline gimmicks, Breaux’s report did find a couple headline writing techniques which led to a consistent increase in click-throughs.
Headlines that used demonstrative adjectives like “this,” “that” and “those” showed significantly higher click through rates and a higher propensity for outperforming other headlines. Demonstrative adjectives can be used multiple ways, such as creating interest with a vague headline like, “You’ve never tasted anything like this.” They can also be used more directly in a call to the reader to seek more time-sensitive information, such as “GOP Debate This Evening.”
Another interesting finding was that long headlines consistently received a higher click-through rate than shorter headlines, something that seems counterintuitive in the era of Twitter and divided attention spans. Breaux agrees, and notes that editors may have actually gone too far in pushing headlines so short they lack enough details to draw readers into a story.
One thing the report didn’t cover was the use of direct quotes as a headline, something I’ve seen more and more on websites. Breaux examined the data and found that headlines that contain direct quotes were 14 percent more likely to win a headline test.
“Speculating a bit into the reasons we see significance, perhaps use of quotes lends a gravitas to what’s being said,” Breaux said. “It may also be a curiosity-driving device. What was the context of the quote? What else did they say?”
Despite these exceptions, Breaux’s data is clear that the best way to write successful digital headlines is to carefully catalogue what works and what doesn’t on your particular publication, and always make sure headlines accurately match the content of the stories they go with.
“Blindly following guidelines can lead to copy that sounds cliché at best, and actively off-putting at worst,” Breaux said. “Effective headline writing can make quite a difference in the success of your content—after all, readers have to get to the actual articles somehow—so it can be expensive to get wrong.”
Rob Tornoe is a cartoonist and columnist for Editor and Publisher, where he writes about trends in digital media. He is also a digital editor for Philly.com. Reach him at firstname.lastname@example.org.