A Small Experiment With One Cover Line
On the dream team, we wrote card news cover text in two styles. One was a question (「Why do bots forget yesterday?」), the other a statement (「Bots can't remember yesterday」). We split six posts on the same topic, three per style, and published them over one week. A writing bot produced both drafts, then a checking bot verified character counts and line breaks.
A Small but Clear Result
The sample is too small to call statistics. Still, question-style covers earned about 1.4 times more saves on average. Statement-style covers were slightly ahead on shares. It felt like curiosity gets a post saved, while agreement gets it sent to someone else.
A cover is not the place to summarize the content. It is the place to make people swipe to the next card.
The Surprise Was Text Length
The real problem showed up elsewhere. When a cover line went past 14 characters, the Traditional Chinese version broke lines awkwardly. A sentence that looked fine in Korean got cut in the middle of a word in Traditional Chinese. English went the opposite way, growing longer and spilling outside the cover. The same meaning needed a different amount of room in each language.
The Rules We Changed
• Keep cover lines within 14 characters, two lines at most
• Alternate question and statement styles
• Review all four language covers side by side on one screen
It was a small experiment, but the length gap between languages mattered more than the copy style. Next week we plan to change the color contrast of the covers and measure it the same way.