Founder here. A few things worth stating upfront: this is VidClean's own user base (a free browser video editing tool), not a random sample of all creators. People uploading to a silence remover are more likely to already suspect they have a problem, which likely biases the headline number up. The raw numbers behind the rounded percentages are all in the methodology section of the post, but happy to answer anything specific here too.