Would you click on an article with the following headline? In Andriivka the Russians are gone again: ‘They searched for ‘Nazis’ as far as the chicken coop.’
Or would you rather click on the next article? In Andriivka, the Russians are gone again: ‘To protect her from rape, I smeared her with cow dung’.
A headline test on nrc.nl gives the answer: you click on the latter more often. No less than 10.3 percent of the readers who were presented with the second headline on the homepage, then clicked through to the whole story. In the first cup, it was only 5.8 percent. The piece received another headline in the paper newspaper: Look there she goes. She fell prey to the Russians.
And this was just one of dozens of micro-experiments with headlines that take place every day on NRC† “We aim to test the headlines of all articles that appear at the top of the site,” says Marie-Louise Schonewille, chief editor-in-chief and responsible for the site’s layout. And it doesn’t stop at headlines: it will soon be possible to experiment with photos on the homepage.
This form of testing is called ‘A/B test’. An article is presented to different readers with different headlines (often referred to as ‘titles’ by readers). Some of the users will see headline A, others will see headline B. The headline that generates the most clicks per view ‘wins’ and becomes the headline that all readers will see afterwards. Your collective reading and clicking behavior of readers are therefore (anonymously) analyzed for this test. NRC uses Chartbeat analysis platform for this.
What does all that testing yield?
Not blissful
First of all, more visitors. Chartbeat estimates that last week, thanks to tested headlines, there were 25 percent more clicks on articles. In about half the cases, the ‘new’ headline was better than the initial headline the piece came online with.
But clicks are not the cure, says Sophie van Oostvoorn, who analyzes reading behavior from the reader’s desk. Chartbeat also records ‘quality clicks’. A click counts as a quality click if a reader spends at least fifteen seconds reading the piece after clicking. If a roaring headline creates expectations that the article doesn’t live up to, a reader is quickly gone. Van Oostvoorn is therefore not afraid of proliferation clickbait† “As editors, we are reluctant to do that anyway.”
Don’t heads drift more quickly towards emotion? Take the headline from Andriivka – rape beats the hen house. Both passages appeared in the report, Van Oostvoorn argues. “A headline should always be in the spirit of the piece.”
In any case, testing delivers Others heads up. You could say: better headlines. Pointed headlines quickly kill online, Schonewille says. That is mainly a matter of context. “A newspaper page offers a lot of information at a glance. Is a piece big or small? Is it a report or interview? But on the site, all the pieces initially look the same: a photo with a headline. It comes down to clarifying: what is this about?”
Errors stand out
The most important lesson for an online headline is: be specific. The digital reader does not always click on cryptic or poetic headlines. And articles omitted in headlines for the sake of space can make an online headline more readable.
And so headlines online sometimes get longer. A report that was in the newspaper under the headline Digging as an archaeologist in your own backyard (68 percent quality clicks) lost out to With trowels, sieves and measuring tapes, the residents of Schijndel in Brabant search for the origin of their own village (79 percent quality clicks).
And so every intuition, every wisdom, can be tested and examined. Are people more likely to click on a headline that talks about the SPD or ‘German Social Democrats’? (The latter.) Does it help if an interviewee’s name is in the headline? (Yes.) Does a one-word headline work with a column? (Sometimes.)
Making a new headline also means that new mistakes can be made. “Like anything you do twenty times a day, sometimes things go wrong,” says Schonewille. “Last weekend we called hydrogen in a winning cup an energy source. Readers or editors point this out to us and that is rectified.” But mistakes are quickly noticeable, in a headline. Van Oostvoorn: “We often receive complaints from the editors than compliments.”
Even the newspaper reader cannot escape the test. Schonewille: “Because we increasingly publish first on the web, winning headlines also end up in the newspaper.” And why not? “The newspaper reader also benefits from a clear and clear headline.”
What else can be tested? “I would like to experiment with the position of theme blocks,” says Schonewille. “But I think the opening of the site is inappropriate. It contains the article that we find most important at that moment, regardless of what the reader clicks on.”
A version of this article also appeared in the newspaper of April 30, 2022
#Opinion #quality #clicks #headline