Comment fonctionne le tracking cross-device
Advertisers want one profile for your phone, laptop, and TV. They combine logins, household network signals, and purchased data to link devices — deterministically when possible, probabilistically when not.
Lien déterministe (haute confiance)
When you sign in with Google, Apple, Meta, or Amazon on multiple devices, the platform can merge activity under one account ID. The same applies to streaming apps, shopping accounts, and work SSO.
Email addresses and phone numbers uploaded by advertisers create custom audiences that match across devices when you log in with that contact info.
Lien probabiliste et foyer
Without a shared login, data brokers and ad graphs guess that devices belong together using shared IP addresses, similar location patterns, and overlapping app installs. Everyone on one Wi‑Fi may share similar ad categories.
Connected TV (CTV) ad IDs can be linked to mobile IDs when the same streaming account is used on a phone and smart TV.
Signaux cross-device courants
- Logged-in accounts — Google, Apple, Meta, Amazon, Microsoft
- Household IP — same router, same outbound address
- Hashed email / phone — CRM uploads matched at login
- Data clean rooms — brands match audiences without raw PII exchange
Ce que notre outil peut montrer
We show IP-based location and ASN, which contribute to household-level profiling. We cannot see your logged-in accounts on other sites or a platform's identity graph.
- Compare IP signals with VPN guidance.
- Review social tracking for platform logins.
Vérifier IP et signaux navigateur
Vérifier IP et signaux navigateurQuestions fréquentes
Does clearing cookies unlink my devices?
Not if you stay logged into the same accounts on each device. Logins are stronger than cookies for cross-device graphs.
Can a VPN stop cross-device tracking?
A VPN changes your IP, which can disrupt household-level guesses, but logged-in accounts and mobile ad IDs still link devices.
What are data clean rooms?
Restricted environments where two companies compare hashed audience lists without sharing raw customer files directly.