Part 5/9:
However, upon closer examination, Ryan and others discovered inconsistencies in Allen’s data. As she clarified on Twitter, her initial analysis was based on flawed data due to issues with her scraping tool, which led to repeated entries that did not accurately reflect real user reviews.
Ryan meticulously checked multiple reviews that appeared identical and found that they were actually from different users with different profiles. For example, a review that started with "man oh man" was duplicated across multiple profiles, but upon inspection, these profiles displayed different review text and scores. This evidence demonstrated that the supposed fake reviews were an artifact of the scraping process rather than genuine spam.