Summary
Across a four-month scrape of the Meta Ad Library (Facebook and Instagram), we identified 36,725 advertisements promoting so-called ‘ghost stores’ in Austria. These shops disguise themselves as local, family-run retailers, often selling fashion or jewelry. In reality, consumers receive poor-quality goods or nothing at all, while consumer rights are systematically disregarded.
We identified keywords associated with advertisements for ghost stores. Through an iterative process, additional phrasing patterns were identified, run through our crawler connected to the Meta Ad Library API, incorporated into the dataset, and refined. This approach resulted in a validated data set for targeted ads between January 1st and April 30th 2025, which formed the basis for further analysis.
The ads cluster around three recurring manipulation patterns:
- emotional appeals through narratives about shop closure
- Promises of high quality and exceptional service
- Emphasis on regional identity
Our findings point to structural weaknesses in Meta's enforcement and response processes, despite their obligations under the Digital Services Act. Even when we reported pages as a Trusted Flagger, they often remained online for weeks. In one case, a reported page was removed, but replacement accounts quickly reappeared and continued running similar advertisements.
Recommendations
Detect and remove recurring scam structures: Meta should implement systematic detection of recurring manipulation patterns, rather than relying on individual, one-off ad takedowns.
Prevent repeat offenders from returning: Meta should ensure that once a page is removed for running fraudulent ads, replacement accounts cannot reappear and continue the same activity under a new identity.
Speed up Trusted Flagger response times: Meta should process Trusted Flagger reports faster and more reliably; reported pages in this study often remained online for weeks despite being flagged.
AdGuardians — funded by netidee.
The research was conducted as part of the “AdGuardians” project, which was made possible by netidee funding from the Internet Foundation. In cooperation with the Austrian Federal Economic Chamber (WKO).
