Preview|META 2Q25:Product Innovation Driving Potential Revenue Acceleration-Part2
META AMM, Super Intelligence, Hyperion/Prometheus/Titan, TikTok, Model Update
We produce META previews each quarter. Last quarter, we successfully backed a contrarian thesis and identified META's significant beat opportunity in 1Q25.
This quarter, we remain excited and have strengthened our confidence in META's revenue growth for both the current quarter and second half after investing additional research resources. While CAPEX investments may lead to gross margin decline next year, considering AI is the most important growth theme for the next decade, we believe the market will be more forgiving of margin pressure than during META's large-scale Reality Lab investments, especially with more revenue upside potential.
We recently spoke with 15 experts and have updates on several key topics:
Overall ad revenue trends for 2Q25 and 2025 expectations
Tariff impact
TEMU/SHEIN impact
Major product developments
Recommendation algorithm progress
META AMM status
META AI progress
CAPEX updates
TikTok growth trends
In this Part 2 report, we primarily discuss points 6-9.
META AMM Progress
Feature Introduction and Review
AMM is an advanced data permission protocol META offers to advertisers. After signing, META returns detailed original ad attribution data (such as campaign names, creative names, device IDs, click times, etc.) to advertisers through third-party MMPs (like Adjust, AppsFlyer).
However, since 2021, AMM was largely deactivated due to tightened privacy policies (GDPR/ATT), putting advertisers' attribution analysis into a "black box" period.
To address the gap left by AMM's deactivation, META introduced MIR (Android) and AEM (iOS), which use SDKs to insert ad information during the installation process to enable attribution capabilities.
But MIR and AEM have significant disadvantages compared to AMM, including limited fields and non-standardized data structures.
META reactivated AMM on June 17th. With reactivated AMM, advertisers can now access fields including: device_id, campaign_name, ad_name, creative_name, click_time, impression_time, event_name, revenue, and network_name.
We researched AMM usage with multiple ad agencies, DSP companies, and advertising clients.
Expert Feedback
China Expert A:
Previously, advertisers couldn't track and measure many data points due to privacy restrictions. Now with META AMM, user long-cycle funnel data is clearer, and each touchpoint's completeness has improved.
In the testing phase, gaming clients saw an 11% decrease in CAC and a 27% increase in 30-day retention. Additionally, combining with Advantage+ automated bidding significantly improved data efficiency. Based on China market experience, when CVR and conversion performance improve by 10-20%, ad budgets typically increase by 5-10%, though this positive relationship may be affected by rising acquisition costs.
In 2Q25, META AMM's budget impact remains very limited (less than 1%), mainly due to the late launch cycle and clients still being in the learning phase. 3Q25-4Q25 may see a 3-4% budget increase.
Uncertain whether this will impact GOOG and APP budgets, as ad platforms consistently improve efficiency annually.
Keep reading with a 7-day free trial
Subscribe to FundamentalBottom to keep reading this post and get 7 days of free access to the full post archives.