Mobile Advertising ID (MAID)

A mobile advertising ID (MAID) is a unique, resettable string that a smartphone gives to apps for advertising purposes. On Apple's iOS it is the Identifier for Advertisers, or IDFA; on Google's Android it is the Google Advertising ID, or GAID, also written AAID. The idea was to give advertisers a stable-but-controllable handle on a device without using permanent hardware identifiers like a serial number or a phone's MAC address, which cannot be changed and would allow unstoppable tracking. A MAID is deliberately different: it is tied to advertising rather than identity, it does not contain a name, and the user can reset it at any time to break the link to past activity, or switch it off. With a MAID, an ad network can recognize the same device across different apps to target ads, cap frequency, and, crucially, attribute an install or purchase back to the ad that caused it.

The mobile advertising ID became the backbone of app marketing measurement, because it let separate apps and ad networks agree on which device did what, which is what makes cross-app targeting and install attribution possible. That backbone has been deliberately weakened. In 2021, Apple's App Tracking Transparency framework made the IDFA available only if the user grants an app explicit permission to track, and a large share of users decline, so the IDFA is now missing for most iOS users. Google has pursued its own Privacy Sandbox effort on Android to reduce reliance on cross-app identifiers like the GAID over time. The direction of travel is clear: the era of freely available device advertising IDs is ending, and measurement is shifting toward consented, aggregated, and privacy-preserving methods. Treating the MAID as an always-available identifier is now a mistake.

A mobile advertising ID is often compared to the web browser cookie, and the comparison helps as long as you note the differences. Both are identifiers that let advertisers recognize a user across sites or apps, target ads, and measure results, and both have come under the same privacy pressure. But they live in different places and behave differently. A cookie is a small file stored by a web browser, scoped to that browser, and third-party cookies in particular are being phased down across the web. A MAID is provided by the mobile operating system itself, shared across the native apps on the device, and governed by the platform's rules rather than the browser's. So a cookie tracks a browser on the web, while a MAID tracks a device across its apps.

The distinction matters because the two identifiers are disappearing along parallel but separate tracks, and each requires its own response. On the web, the answer to third-party-cookie loss has been first-party data, consented identifiers, and aggregated measurement. In apps, the answer to MAID restriction has been platform-provided attribution frameworks — Apple's SKAdNetwork and its successor, and privacy-preserving APIs — plus first-party relationships and modeled measurement. An advertiser who fixes only the cookie problem and assumes app measurement is unaffected, or the reverse, will be caught out, because the device MAID and the browser cookie are governed by different companies with different rules and timelines. They are cousins in function and in decline, not the same identifier, and a durable measurement strategy has to address both worlds on their own terms.

Working well in the MAID's twilight means building measurement that does not assume a device identifier is there. Because so many users now withhold the IDFA and platforms are restricting cross-app IDs, lean on consented first-party data — the relationships and information users share with you directly — and on the platforms' own privacy-preserving attribution tools rather than raw device IDs. Use aggregated and modeled measurement, and validate channels with incrementality tests that estimate real lift instead of leaning solely on last-touch attribution that a missing MAID can no longer support. Respect consent as a feature, not a hurdle, since the identifiers are restricted precisely to protect users. And keep your measurement partners and SDKs current, because the frameworks that replace the MAID are still evolving and change frequently.

The failures come from clinging to a fading identifier. Some teams still assume every device offers a usable MAID and build attribution that quietly breaks for the majority of users who opt out. Others treat the loss of the IDFA as the whole story and forget that the browser cookie is disappearing on a separate track, leaving half their measurement exposed. Some ignore the platforms' privacy-preserving frameworks and consented first-party data, then wonder why their numbers no longer reconcile. And a few chase workarounds like device fingerprinting that platforms prohibit and may penalize. The discipline is to accept that freely available device advertising IDs are ending, shift toward consented and aggregated measurement, keep tooling up to date, and judge channels by incremental results rather than by a device ID that many users no longer share.

Mobile advertising ID describes its function plainly — an identifier used for advertising on mobile devices — realized as Apple's IDFA in 2012 and Google's advertising ID in 2013.

Etymology: source.

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Areas of marketing where mobile advertising id (maid) is a core concern: