Monday, 4 November 2013
Linda Lai and Audun Farbrot used a real science communication Twitter feed that had 6,350 followers at the time of the study. Real stories were tweeted to these followers twice, an hour apart. The first tweet used a statement headline, such as "Power corrupts". The second tweet, referring to the same story, was phrased as a question that was either self-referencing, as in "Is your boss intoxicated by power?" or non-self-referencing, as in "Are bosses intoxicated by power?"
Lai and Farbrot found that self-referencing question headlines were clicked on average 175 per cent more often than statement headlines (this advantage dropped to 150 per cent for non-self-referencing question headlines). The difference in clicks for question and statement headlines was statistically significant, but the difference between the self-referencing and non-self-referencing headlines was not.
A follow-up study was similar but was conducted via the Norwegian equivalent of Ebay, known as Finn.no. Lai and Farbrot posted adverts for an iPhone, a couch, a TV and a washing machine using either statement headlines or question headlines (self-referencing or not), such as: "For sale: Black iPhone4 16GB"; "Anyone need a new iPhone4?"; or "Is this your new iPhone4?"
Overall, across the four products, non-self-referencing question headlines were clicked on 137 per cent more often on average than statement headlines; this rose to 257 per cent more often for self-referencing question headlines. This time the difference between the two types of question headline was statistically significant. This overall benefit of question headlines was observed despite one anomaly that the researchers were unable to explain - question headlines for washing machines actually led to fewer clicks than statement headlines.
Lai and Farbrot cautioned that they've only investigated the power of question headlines in a limited context. Another potential criticism of their work is surely that some of the question and statement headlines may have differed in other ways besides their quizzical status. In the previous example about bosses, for example, the question headlines were longer and referenced the concept of "intoxication" whereas the statement headline did not. Assuming questions really do provoke more clicks than statements do, another weakness of this paper is that it doesn't tell us anything about why this is the case.
These issues aside, the clear take-out from this research is that you should phrase your headlines as questions, especially self-referencing ones, if you want to attract more clicks. "The combined strategy [of question headlines and self-referencing] seems to represent a useful tool for practitioners in attracting readers to their Internet-based communications," the researchers said. However, an issue they don't address is what happens if headline writers heed this message and adopt question headlines universally. Perhaps then statement headlines would appear more original and distinctive and attract more clicks. What do you think?
Linda Lai, and Audun Farbrot (2013). What makes you click? The effect of question headlines on readership in computer-mediated communication. Social Influence DOI: 10.1080/15534510.2013.847859
Post written by Christian Jarrett (@psych_writer) for the BPS Research Digest.