Myth: high Limit-Ad-Tracking / opted-out share is a fraud signal

Myth: high Limit-Ad-Tracking / opted-out share is a fraud signal

A filter floating around fraud forums: a publisher with lots of LAT or ATT-opted-out users is suspicious. You'll blacklist your cleanest sources.

LAT and ATT opt-out rates vary by region, OS version, and audience sophistication — privacy-savvy users opt out. A high opt-out share is a demographic fact, not a fraud tell. Real fraud often does the opposite: faked traffic frequently presents consent to look attributable.

✓ Anomalous 100%-opted-out from one sub-publisher can be a flag
✓ Worth correlating with other signals
✗ Opt-out rate alone tracks privacy behavior, not fraud
✗ Blacklisting high-LAT sources cuts premium privacy-conscious audiences

Weight opt-out rate near zero in your fraud score. Lean on CTIT shape, install velocity anomalies, and post-install engagement instead.

Verdict: use opt-out share as weak context at best, skip it as a primary fraud signal — it profiles privacy, not fraud.
Best for: analysts who built fraud rules on opt-out percentages.