Media manipulation is more than bot traffic.
The framework separates Content / Platform manipulation from Identity manipulation and distinguishes directly ad-relevant mechanisms from mechanisms that are not directly advertising-specific.
CONTENT / PLATFORM
AD RELEVANT
- Fake Content
- Content Scraping
- Fake News & Disinformation
- Fake Sites / Spoofing
- Dark Pooling
- Auto Reload & Refresh
- Ad Stacking
- Tag Manipulation
- Sourced Traffic / Ad Laundering
- Invisible Ads / Blank Codes
- Pop-unders
- MFA / Made-for-AdvertisingSUPPLY PATTERN
NON-AD RELEVANT
- Deepfakes (Video / Audio)
- Shallow Fakes (Image / Video / Audio)
- Fake News & Disinformation
IDENTITY
AD RELEVANT
- Fake Traffic
- Fake Profiles
- Bots
- Fake Devices / Fake Apps
- Background Ad Loads
- Device ID Rotation
- Random Device ID
- Retargeting Fraud / Cookie Harvesting
- Incentivized Views
- Residential Proxies
- Malware / Adware
- Fake Attribution
- Click Injection
NON-AD RELEVANT
- Deepfakes (Video / Audio)
- Shallow Fakes (Image / Video / Audio)
- Bots
- Fake Profiles
Working model: categories can overlap. MFA is a supply/content/business-model pattern rather than one single fraud mechanism. The taxonomy is not exhaustive.
Two layers of manipulation
Content / Platform manipulation concerns content, sites, delivery logic or technical placement mechanics. Identity manipulation concerns traffic, profiles, devices, apps or other identity signals that can simulate human or legitimate activity.
Ad-relevant is not an exclusive category
A mechanism can directly affect advertising while also occurring outside advertising. Bots, fake profiles or deepfakes therefore cannot always be confined to one column. The taxonomy primarily describes the media and audit context rather than a universal ontology.
MFA needs separate context
Made-for-Advertising is not a single fraud mechanism such as click injection or ad stacking. MFA is better understood as a supply, content and business-model pattern that can create quality, transparency and manipulation risks depending on its implementation. The framework therefore marks MFA as a context or supply pattern.
Why the taxonomy matters for audits
A single fraud rate says little about the manipulation mechanism behind a finding. An audit should distinguish whether anomalies originate in content, platform logic, traffic, device identity, attribution or the supply chain, and which independent signals support that interpretation.
No single signal proves manipulation
Media Forensics evaluates technical, behavioral and contextual evidence together. An anomaly is initially a reason to investigate. A more defensible classification and proportionate action require coherence across multiple independent evidence points.
This article describes principles and questions. Specific fraud, bot or quality rates cannot be transferred directly to individual campaigns without understanding methodology, scope and data basis.