Thanks, Anya. Considering we are in the age of Big Data and AI, and given the vast amount of very specific data the platform has about every reader-writer interaction, the mechanism they currently have in place seems woefully, embarrassingly inadequate.

And even undermines the Boost programme itself, as on Day 1 it seemingly just spurts a boosted piece out to completely random figures, often generating a low read ratio, which then hampers further distribution.

It's 20th-century data science for a 21st-century platform.