Download This PaperLeveraging Public Traces to Monitor Fake-review Campaigns
30 Pages Posted: 2 Jun 2026 Last revised: 2 Jun 2026
Date Written: June 02, 2026
Abstract
The dominant narrative in fake review detection is that it is primarily a platform-driven function.
While digital platforms can employ internal private data to spot fake reviews, such data is
usually unavailable to external parties or regulators to audit fake review claims. Compounding the
problem is a recent trend in recruited fake review campaigns on social media, that employ real
reviews buying real products, whose reviews appear to be more authentic than fake. We present
an alternate method of fake review detection that employs public data to recover 92.2% of independently observed recruited reviewers. It also identifies the right products: among the top 100
suspicious review bursts surfaced by the method, 87% correspond to products independently observed in recruitment campaigns in the Facebook broker groups. Scaled to the 2023 Amazon review corpus, our method suggests that about one in eight verified-purchase 5-star reviews and nearly one in five 5-star reviews sit within high-evidence coordinated-recruitment structures recoverable from public traces. We showcase that fake-review monitoring does not have to remain locked inside platform black boxes. Regulators, journalists, watchdogs, and consumers can use public traces and our simple method to identify suspicious reviewers, products, and sellers and hold platforms accountable.
Keywords: recruited reviewers, e-commerce, networks, policy for fake reviews, fake review monitoring
Suggested Citation: Suggested Citation
