Quantcast
 

On Quiz: Can You Tell What Makes a Good Tweet? (Corrected)

The NYT examples all have the date and time attached, which not all of these examples do. If you examine the differences between the tweets, and the algorithm's predictions, the criteria they are using to evaluate the potential success will become pretty clear. (Still, a mildly psychopathic marketing human can do much better: 20/25, when the algorithm only got 13/25. https://twitter.com/Partorio/status/484222836253671424)

The titles of these articles mislead, however, since the texts nowhere deal with evaluating the quality of the tweet, but its relative marketabilty. There was only one example in the NYT, from British comedian David Mitchell, which was even functionally literate, while the others seen were only slightly less horrific than the ones above by dint of containing some glimpse of information, or at least a pointer to some original content.

The concept of limited characters and a rapidly moving format isn't itself outrageous, it's the marketers abuse of the medium, and the public's willingness to see themselves as products to be marketed, that degrade the platform.

Posted on July 2, 2014 at 4:19 pm 0