When ATT arrived with iOS 14.5 in April 2021, publishers dependent on IDFA saw iOS CPMs fall sharply. Before ATT, IDFA gave DSPs a persistent cross-app identifier for retargeting, cross-app frequency capping, and lookalike modeling. Buyers paid more for identifiable users. Publishers that rebuilt their mediation around available contextual signals recovered much of the loss. The drop was real.
More than four years later, the outcome is less uniform. Publishers that configured mediation to transmit available contextual signals correctly have recovered most of the gap; those that did nothing have not. Post-ATT iOS monetization comes down to identifying the remaining signals and passing them correctly.
What IDFA Enabled
IDFA supported several advertising uses, and the shift to opt-in access did not reduce their value equally.
Retargeting reconnects advertisers with people who previously used their app or website, and it depended heavily on IDFA. Without a persistent cross-app identifier, scaled iOS retargeting became difficult. Publisher CPMs took the biggest hit here because retargeting campaigns generally bid most for known users.
Cross-app frequency capping stopped users from seeing one ad repeatedly across publishers and also relied on IDFA. Without it, advertisers may over-serve users or under-serve them cautiously. Both reduce publisher eCPMs: over-serving harms ad quality, while under-serving lowers bid density.
Lookalike targeting builds audiences resembling known high-value users. Without the IDFA graph, precision declined, but contextual and probabilistic methods adapted to this use case more successfully than to retargeting.
Targeting Signals Still Available
Contextual, device-level, and geographic signals remained after ATT. When your mediation SDK passes them correctly to demand partners, advertisers can target with meaningfully more information than in a blind auction, supporting higher CPMs.
App category and content classification: The App Store category, such as games, utilities, or news, plus genre tags in the bid request, tells DSPs the user's context. A finance app user receiving a financial product ad is a contextual match worth a premium, even without user identification.
Geography: Country and region targeting do not require IDFA. Location is valuable to advertisers focused on particular markets. Publishers concentrated in high-CPM geos can regain a substantial share of the IDFA-era eCPM premium on that traffic.
Device type and OS version: iPhone or iPad, current iOS version, and device model remain available. Advertisers concerned with the rendering environment use these signals. Some advertiser categories bid more for recent iOS versions on high-end devices.
SKAdNetwork attribution: Apple's privacy-preserving framework measures install campaigns without IDFA. Correct SKAdNetwork reporting helps advertisers complete measurement, increasing confidence when bidding on your inventory. The effect is downstream, but real.
Configure Mediation for These Signals
The signals are available, but many publishers skip or misconfigure the required demand-partner settings. Adrop's SDK includes available contextual signals in each bid request by default. Placement configuration still needs verification.
App Store category and bundle ID: verify that the bundle ID in Adrop's dashboard exactly matches the App Store bundle ID. A mismatch prevents demand partners from mapping the request to a known category and lowers its value. This small setting can materially affect eCPM.
Geo fields: check that the mediation SDK sends the user's country code in bid requests. Some GDPR integration patterns unintentionally strip or mask geo fields for IDFA-opt-out users worldwide. With in-app GDPR consent management, confirm geo remains available for non-EU traffic and for EU users who consented.
Content URL and page context: content apps such as news, video, and publishing can pass a URL or content category in the bid request for page or article-level contextual targeting. This matters more for banners than rewarded video, but remains relevant in mixed-format setups.
ATT Prompt Effects on iOS Fill
Whether improving the ATT prompt to raise opt-in rates is worthwhile depends on your user base and current opt-in rate.
ATT opt-in differs considerably by app category. Games generally have lower rates than productivity or utility apps because users may be more privacy-aware or see less value in the exchange. Timing also matters: users prompted after receiving app value opt in more often than users prompted at first launch.
Maximizing opt-in is not necessarily the primary lever. For most publishers, contextual and geo targeting plus SKAdNetwork integration closes the CPM gap more reliably than moving opt-in from 20% to 30%. Both contribute, but mediation configuration is directly controllable and remains consistent regardless of user behavior.
Current Publisher Position
Publishers investing in contextual signal configuration, geo-aware floor management, and multi-network mediation are usually within striking distance of pre-ATT eCPM levels on iOS. Recovery is uneven: retargeting-heavy categories such as gaming and e-commerce still have thinner iOS inventory than before 2021. Utility apps, content apps, and casual games have seen relevant buyers rebuild around contextual signals.
The publisher still underperforming on iOS in 2026 often configured mediation in 2020, never revised it for ATT, and now sees iOS CPMs below Android. This configuration gap is real and fixable. A signal audit is the practical remedy, rather than waiting for the market to recover independently.
[/BODY-HTML]