In accordance with MRC guidelines, domain mismatchesare considered a potential indicator of SIVT under the category of Domain and App misrepresentation: App ID spoofing, domain laundering and falsified domain / site location. However, there’s a significant segment of domain mismatches that can be tied to valid traffic. HUMAN defines indicators to classify mismatches as either invalid or valid traffic. These are based on post-bid detection.
The mismatching domains host websites that are similar in content, branding, structure, and audience. These mismatches aren’t flagged as IVT, but can still be reviewed using the Domain Mismatch flag in the HUMAN Dashboard.
No discernible difference between the websites hosted at the intended and detected domains in terms of content, audience, structure, or localization.
Mismatches caused by a mobile version of a website are also considered valid traffic.
For example: m.example.com vs example.com or amp.example.com vs example.com are considered benign mismatches.
Both domains point to the same website (for example, with a redirect) with the same content, branding, structure, and audience, and there is evidence that the domain spoofing is likely non-malicious.
For example: Intended domain is news-site.com and detected domain is news-site.co.uk, and news-site.com redirects to news.site.co.uk.
Traffic is loaded within a Google website but the user is still being presented with an iframe of the intended domain.
For example: Intended domain is new-site[.]com but detected domain is news[.]google[.]com. However, we can detect that new-site[.]com is loaded in an iframe on top of Google news.
Domain mismatches are flagged as IVT when the domains are considered materially different based on one of the indicators described below. These mismatches are tied to the category SIVT > False Representation > Domain Spoofing.
Mismatches can be reviewed using the Domain Mismatch flag, and they’re also identified under the Domain Spoofing IVT subcategory in the HUMAN Dashboard.
news-site[.]com vs. gaming-site[.]com.news-site[.]com vs. news-site[.]netnews-site[.]com vs. news-site[.]hknews-site.com and detected domain is news-site.co.uk, and news-site.com redirects to news.site.co.uk.health-news-site[.]com and detected domain is a subsection of that site that lists doctors doctors[.]health-news-site[.]com. This has a materially different page structure.dictionary-site[.]com with word etymologies that has 2-3 ads on the sides and detected domain is a subsection of that site word-game[.]dictionary-site[.]com with a word game that has auto-playing video ads and 10+ ads per page.news-site[.]com and detected domain is sports.news-site.com.sports.news-site[.]com and detected domain is news-site.com.news-site[.]com and detected domain is uk[.]new-site[.]com.
These mismatches aren’t flagged as IVT when the intended domain is already region-specific. For example: Intended domain is brazilian-soccer[.]com and detected domain is brazilian-soccer[.]br.
Detected domain is related to an Ad Verification provider or other known sandboxed environment.
Intended domain is shown in pre-bid as a domain typically associated with ad loading or ad tech infrastructure, unlikely to be the actual domain deliberately being transacted. Detected domain is something else.
For example: Intended domain is googlesyndication[.]com and detected domain is something else.