A survey result that the viewer starts to stop viewing when the movie playback time is delayed by 2 seconds
The time that people can tolerate until the web page is displayed in 2006 is "Up to 4 seconds"The survey results say thatAkamaiAlthough it was made public by the survey, in the survey it turned out that for the movie people started to close off in 2 seconds.
(PDF)The Impact of Video Stream Quality on Viewer Behavior
Online viewers start leaving if video does not play in 2 seconds, says study - Tech News and Analysis
This survey was conducted by Professor Ramesh Sitaraman, a member of Akamai Technology and studying computer science at the University of Massachusetts, based on 23 million movies and the data of 6.7 million unique users who reproduced it It was.
As a result of the investigation, the faster the Internet line became, the more patience was lost, and when it could not be read within 2 seconds, it turned out that some people began to stop watching the movie. At this time, people seem to stop seeing by 6% each time the reading is delayed by 2 seconds. The following graph shows two types of movies, one with long playback time and one with short playback time, but for those with short playback time, 20% viewers stopped watching when reading was delayed by 5 seconds I will. On the other hand, as the time of the movie itself becomes longer, a person seems to be patient.
In 2006, "I can wait until the website is displayed is up to 4 seconds"4 second ruleAlthough it was announced, this survey resulted in supplementing that "With respect to movies people become even more persistent" in this 4 second rule. By the way, different results have come out depending on the line you are using. People become the most patient in accessing from mobile devices, and in the case of optical fibers, as long as playback of 10 seconds is delayed, it seems that more than 60% of people will stop looking.
These research results are not surprising, but it seems to be helpful for companies with a video distribution platform, for example.
in Note, Posted by logq_fa