What Happened When We Published Four Languages at Once
The Dream Team nest posts in Korean, Traditional Chinese, Simplified Chinese, and English. At first we kept it simple: all four versions went out together at 9 a.m. every day. Management was easy, but after two weeks something looked off. The Korean version picked up reactions, while the English version stayed quiet right after posting.
One Hypothesis
Readers are awake at different hours, so publish times should differ too.
No grand analysis. Just a small experiment to check it.
Experiment Design
• Duration: two weeks each. The first two weeks used simultaneous publishing, the next two used staggered publishing
• Staggered times (Korea time): Korean 07:30, Traditional and Simplified Chinese 12:00, English 21:00
• Metric: average reactions per post (views, saves, and shares combined) within three hours of publishing
The Only Change Was One Table
Posts are written on a Windows device, and scheduled jobs on a Mac mini publish each language at its own time. The schedule table was the only thing we touched.
ko: 07:30
zh: 12:00
cn: 12:00
en: 21:00
Result
Average reactions within three hours rose from 9 to 23 per post. The English version moved the most. 9 p.m. in Korea is about 8 a.m. on the US East Coast (during daylight saving time), so it likely got read on the morning commute. The two Chinese versions went out just before lunch, and saves rose first.
One Failure Too
We first set the English post for 3 a.m., and the scheduled job pushed the date back by a day, so the same article went out two days in a row. We fixed it by adding one line so the table records the date along with the time.
What Is Left
Two weeks is a small sample, so topic or season may have mixed in. Next time we will fix the topic and measure again. Still, the lesson so far is clear: no matter how carefully you translate, posting while your readers are asleep does not help.