DV uncovers large-scale audio ad fraud scheme

fm-scam-ad-fraud-uncovered-by-dv

DoubleVerify, a software platform for digital media measurement, data, and analytics, today announced the discovery of “FM Scam,” the latest audio fraud scheme to date that was identified and mitigated by DV’s Fraud Lab using a proprietary combination of AI-powered technology and human review.

Audio ad fraud scheme

FM Scam represents the second significant global scheme targeting audio spending, emerging shortly after the DV Fraud Lab protected advertisers from BeatSting, the first major invalid traffic (IVT) scheme aimed at audio traffic.

During the peak of both schemes, the combined financial impact on unprotected advertisers exceeded $1 million in spend monthly.

FM Scam perpetrators falsify audio traffic using dedicated servers. They also spoof a wide range of devices typically used for playing audio content.

This tactic enables fraudsters to blend their invalid traffic with legitimate traffic in attempts to remain undetected.

DV estimates that the fraudsters behind the scheme spoofed over 500,000 devices, including mobile phones, tablets, connected TV (CTV) devices, audio players, smartwatches and, for the first time, even smart speakers.

Fraudsters auction their invalid traffic through supply-side platforms (SSPs), exchanges and ad networks. DV’s impact analysis indicates that FM Scam generates up to 100 million ad requests monthly.

In addition to defrauding advertisers, the fraudsters also are impacting the yield of quality audio sellers by syphoning money away from legitimate audio channels.

“Audio is rapidly becoming a lucrative target for fraudsters,” said Jack Smith, Chief Innovation Officer at DoubleVerify.

“The range of players and devices used to stream audio content sets this environment apart from other media types.

“Fraudsters tend to focus on emerging media, like audio, that attracts advertising investments but lacks comprehensive measurement standards.”

FM Scam further blends in with legitimate traffic by mimicking the behaviour of a typical continuous ad session. Unlike BeatSting, which switches between devices after generating a certain number of impressions, FM Scam creates longer, uninterrupted sessions.

This randomises the traffic pattern to appear more human-like, as opposed to delivering a consistent, fixed amount of impressions across devices.

Traffic Patterns in BeatSting vs. FM Scam

fm-scam-ad-fraud-dv-charts
Source: DoubleVerify

When FM Scam emerged, the Fraud Lab quickly tied IP addresses used by the attack to a CTV scheme first detected in 2019. Some IPs were also associated with various malicious activities, including the spread of malware.

The DV Fraud Lab immediately mitigated FM Scam’s attacks and has been continuously shutting down new variants of this scheme leveraging a proprietary combination of AI-powered technology and human review.

DV’s Fraud Lab — powered by a dedicated team of data scientists, mathematicians and analysts — performs ongoing detection and analysis of new types of digital ad fraud to uncover the latest schemes as they occur.

Through continuous analysis, scenario management and research, the Fraud Lab pinpoints the sites, apps and devices responsible for fraudulent activity and updates protection for DV clients in real time.