Does clicking mean reading? Are clicks enough for the success of a newsroom? 

At the recent INMA Media Subscriptions Summit in Amsterdam, industry experts showcased their research collaboration on headline testing and introduced an AI-powered tool that offers a dynamic approach to tailoring content based on real-time audience needs.

Headline formulation: Negativity drives up clicks

What kind of headlines generate the most clicks? 

According to Luuk Lagerwerf, associate professor at VU University Amsterdam, an A/B testing of 6,000 headlines from national and regional newspapers and news Web sites in The Netherlands showed emotional appeal, negativity, and narrativity can increase click-through rates.

“The use of negativity has the largest positive effect (on clicks)… . But on the other hand, it’s quite ugly to use negativity, which means the negative words and words with a negative sentiment.”

Meanwhile, certain linguistic elements, such as consonance, can create audience fatigue, leading to disengagement.

“If we have a consonance like the word ‘still’ in the headline to emphasise that ‘it’s still going on,’ we actually didn’t expect too much positive effect of this because if it’s still going on, it’s not new, it’s not attractive to click on,” he explained.

Notably, headline effectiveness may vary by brand. In Lagerwerf’s A/B testing, while negativity or personalisation boosted clicks for some newspapers, it had no impact on others.

“Super detail” and loyalty clicks

But do more clicks mean more loyal readers? 

Roy Wassink, insights manager at DPG Media, warned that while negative headlines tend to attract attention to every reader, they don’t necessarily build reader trust.

“If the headline is negative but the first paragraph is positive, which is a mismatch, half of the readers may drop off before finishing the story,” Wassink explained.

He also cautioned against the long-term consequences of clickbait-driven negativity, citing research showing nearly 50% of Dutch readers actively avoid the news, including 22% of who “sometimes” avoid it and 29% who do so “occasionally.”

“Clicks don’t equal engagement,” he emphasised. “And more clicks don’t necessarily mean more readers.” Instead, newsrooms should prioritise loyalty clicks — engagement that extends beyond a fleeting glance.

“If readers stay on the page for at least 10 seconds, they’re more likely to return — and eventually subscribe,” he added.

Wassink also highlighted the importance of seemingly minor details in headlines. “Super detail,” he said, adds extra colour and brings something to this story that makes people click.

“If you have a good story, show people,” Wassink said. “Give them a promise in the headline that it is a good story. If we compare a factual headline with a more story-like headline with details like names, especially those of ordinary people, the story-like headline is much longer, isn’t it? But it can create a sense of identification and thus increase clicks by about 25%. What we say to our editorial teams is: If you can bring in some super detail, do so.”

Wassink mentioned that such “super detail” does not just generate more clicks, it also enhances reader engagement and fosters loyalty. However, he cautioned against the overuse of it: “Of course, you can’t fill an entire Web site with names in every headline.”

Beyond A/B testing: smarter, AI-driven content strategy

Moving beyond traditional A/B testing, Rutger Verhoeven, chief marketing officer at Smartocto, shared details about an AI-powered solution that allows newsrooms to define their own performance metrics beyond clicks and likes.

Smartocto’s AI tool leverages a User Needs Model, providing real-time insights, recommendations, and notifications to writers. This enables dynamic content adjustments based on actual audience behaviour.

The AI tool serves as a support rather than a replacement for editorial judgment, Verhoeven said: “Of course, you are not gonna change what the story is about, but you can change the words, the sentiment, the tone, the language, and the accessibility of the story.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .