AppsFlyer Fall Release 2025: Must-Know Updates
"Today isn't just another product launch. Today we're taking the first step in shaping the future of marketing,"stated AppsFlyer Co-founder and CEO Oren Kaniel.
This landmark update introduces the Modern Marketing Cloud, a platform explicitly built for the AI era and designed to be independent, trusted, and privacy-first. After successfully leading the mobile revolution and navigating the turbulent waters of the privacy revolution (specifically iOS 14 and SKAN), AppsFlyer is now dedicated to helping the mobile community "lead the AI revolution" inside their companies.
This blog opens with what the Fall Release 2025 includes and then moves through its main pillars: measurement upgrades, AI tools, creative automation, and data collaboration. It closes with the broader platform changes to show how all updates connect and impact daily workflows.
What Is AppsFlyer Fall Release 2025?
The Fall Release 2025 is a major product-launch initiative by AppsFlyer, announced around November 18, 2025. It introduces new products and capabilities across multiple solution suites, including measurement, AI, data collaboration, and security.
It represents an evolution of AppsFlyer from a mobile attribution platform to what they describe as a “Modern Marketing Cloud” that unifies measurement, data collaboration, and autonomous workflows.
Key Focus Areas:
Some of the notable updates included in the release:
An “Agentic AI Suite”: Enabling marketers to deploy AI agents, connect via a Model Context Protocol (MCP) to large language models (LLMs) such as ChatGPT, Claude, and others within the AppsFlyer environment.
A “Signal Hub”: A privacy-safe data collaboration layer combining first-party data, partner data, and clean-room technologies to handle signal loss and fragmentation.
Expanded cross-platform measurement: Stitching journeys across mobile, web, desktop, console, and CTV, and linking revenue/events back to campaigns.
Enhanced enterprise security package: SAML 2.0 SSO, SCIM provisioning, multi-token governance, audit logs, IP allow lists, and granular RBAC.
Updated attribution model: Applying real-time AI behavioral analysis to detect fraud, such as click flooding, to produce cleaner attribution data.
Redesigned dashboards (“My Dashboards”): With natural-language queries, an AI assistant embedded, and unified views (Activity, LTV, Cohort, SKAN, etc.).
Creative Management Hub: Centralized storage, scoring, and deployment of creative assets, with AI-powered insights.
These updates highlight the main features of the Fall Release 2025 for the platform. To explain how these elements fit together, the next section provides an in-depth breakdown of each update and the reasoning behind the new features.
Also Read: Mastering the Power of Incrementality in Mobile Games and App Marketing
Detailed View of the Fall Release Updates
With the broader themes established, the release becomes clearer once you look at the pillars that AppsFlyer focused on. The first area centers on measurement and its reworking:
The Future of Measurement
AppsFlyer launched three major products to take measurement confidence to a new level, moving beyond just attribution (the "what") to impact measurement (the "why").
1. Incrementality for User Acquisition (UA)
Goal: To measure the true impact of marketing campaigns by answering the question of what truly drives growth and eliminating external noise (such as seasonality, CRM actions, word of mouth, or competitors).
Mechanism: This product turns incredibly complex models and massive data sets into one simple click. AppsFlyer creates groups of exposed and hold-out groups (which behave exactly the same prior to the experiment), and the difference between them is the lift.
Impact: The product is now available for general availability. Beta results showed that 30% of campaigns brought up to 10x more conversions than originally attributed, while 18% were not incremental at all (not generating new user intent). It reveals the campaign's true influence (the why) alongside attribution (the what).
The next product continues this track by tackling performance distortion and bringing sharper attribution signals.
2. Enhanced Attribution Model
Challenge Addressed:Click flooding, a type of ad fraud that distorts performance and attribution.
Mechanism: Features rebuilt attribution protection layers. It combines real-time AI analysis with flexible per-device attribution logic to ensure every attribution decision reflects authentic user intent.
Impact: Customers adopting this model saw a 100% lift in their growth KPIs because the right sources received credit and budgets were correctly allocated.
Following attribution improvements, the release moves on to address fragmented user journeys and incomplete value calculations.
3. Cross-platform Journeys and LTV
Goal: To break down measurement silos and partial user journeys that create blind spots and incomplete user LTV. It provides the true ROI of campaigns.
Mechanism: The product links all signals from all leading digital platforms (mobile, web, CTV, PC, console) into one unified reality. It focuses on the user, not devices. It stitches together all touch points and measures all in-app events and revenue to calculate the total and complete user LTV and attribute it back to the original acquisition source.
Dashboard Features: Introduces a new grouping method called a product line to group multiple assets that serve the same product (e.g., mobile app, CTV, PC).
Impact: Beta tests showed significant improvements, including up to a 65% increase in attributed revenue (entertainment apps) and up to 48% increase in ROAS.
Future Milestones: The first milestone is launched, with future improvements including a new web attribution model and improvements to PC and CTV attribution coming soon.
Once the measurement foundation is laid out, the next major cluster focuses on AI-driven execution and operational efficiency.
Agentic AI Suite (Autonomous Marketing)
This suite is built on the philosophy of balancing human creativity and strategy with the power of AI, moving marketing toward autonomous operations.
4. AI-Ready Data and Infrastructure
Principle: AI is only as good as your data; otherwise, it is "garbage in, hallucination out".
Data Layer Revamp: The entire data layer (every table, column, query accessible through Data Locker, Push API, or Pull API) has been revamped to ensure data is clean, connected, and crystal clear.
Intelligent Metadata Layer: This is the "secret weapon" providing data about the data. It is the connective tissue that allows AI to understand, query, and act based on the user's data.
With the data layer tuned, the release introduces tools that turn this foundation into direct support for day-to-day decisions.
5. AppsFlyer’s AI Assistant
Functionality: A 24/7 secret partner integrated into the dashboard experience. It enables marketers to ask questions in natural language in multiple languages (English, Korean, Japanese, and Spanish).
Benefits: Provides top campaign data and suggestions for deep dives. It can generate an executive summary of the past week’s performance in seconds, highlighting key changes and suggesting what to look at next.
Building on the assistant, the next capability widens the use of LLMs by connecting them directly to live marketing data.
6. Model Context Protocol (MCP)
Functionality: A platform that connects AppsFlyer data directly to a customer's preferred Large Language Model (LLM) of choice (e.g., ChatGPT, Claude, Gemini).
Mechanism: MCP translates natural language questions into data queries, pulls the secure data, and returns the answer in natural language. It can answer sophisticated questions that once took analysts weeks.
AI-to-AI Collaboration: Announced partnerships with Amplitude and Braze, enabling the full marketing cycle to be ready for action.
To bring these AI components together, the Agent Hub acts as the workspace where different agent types operate.
7. Agent Hub
The new home for marketing agents that guides users to autonomous marketing. Agents are categorized into functional buckets:
Alerting Agents: Watch metrics and flag issues.
Opportunity Agents: Scan data to find new growth avenues (e.g., the Creative Opportunities agent).
Reporting Agents: Slice and dice data into clear stories (e.g., the Weekly Marketing Report agent).
Beyond analytics and agent workflows, the release also addresses the creative process, which often dictates performance outcomes.
8. Creative Management
Creative is the single biggest driver of marketing performance. The existing creative engine analyzes over 300,000 creatives every day and optimizes over $5 billion of spend.
Goal: To transform the creative process into a fully automated, end-to-end engine.
Key Features:
Seamless Sync: Automatically syncing and organizing creatives (e.g., from Google Drive) by app, date, or theme.
AI Pre-flight Scoring: The recommendation engine analyzes each creative pre-flight (before spending a dollar) and evaluates its performance potential based on benchmarks.
One-Click Distribution: Distributing top-scoring creatives directly to networks (like Meta) in the correct versions.
After covering measurement and AI operations, the final major pillar focuses on how data partners can work together without compromising privacy.
Data Collaboration
Data collaboration is the third foundation, designed to activate every available signal and enrich data by partnering with others.
9. Signal Hub
A curated privacy safe data marketplace purpose-built for marketers inside the AppsFlyer data collaboration suite. It is designed to move from siloed datasets to data collaboration and from signal loss to signal richness.
Partnership: Launched globally with Mastercard as the first third-party insights provider.
Mechanism: Brands and data owners bring their audiences, and activation happens inside governed clean rooms. Signals arrive aggregated and anonymized, matched with the user's own data, and flow straight into activation without a single row of personal data ever changing hands.
Benefits: Allows marketers to define custom, privacy-safe audiences (e.g., US men aged 20-45 who spent over $200 with specific merchants) based on competitive intent, spend ranges, and real-world behavior.
Alongside headline updates, the release includes broader platform improvements that reshape everyday team usage.
General Platform Enhancements
The first of these improvements is a refreshed dashboard experience designed to bring scattered views into one frame.
10. The New Dashboard Experience ("My Dashboards")
Goal: To create a smart, simple, and helpful data tool for marketers by unifying previously separate views.
Functionality: Unifies data that previously lived across Activity, LTV, Cohort, SKAN, and SSOT. Users can customize every widget, explore any metric, and compare User Acquisition (UA) with retargeting data in a single space. It is highly customizable and adaptive, allowing users to filter by geo or media source at the dashboard level or within individual widgets.
On the operational side, enterprise teams get a more controlled and transparent security setup.
11. New Enterprise Security Package
Goal: To provide control, visibility, and peace of mind for enterprise teams.
Features: Includes SCIM granular token management, extended audit log, and IP allow list.
The entire release is structured like a Jenga tower: confidence in measurement is the foundation, enabling the building blocks of autonomous marketing, which ultimately leads to the upper layer of data collaboration.
Also Read: Guide to Understanding Ad Attribution Models
Conclusion
AppsFlyer’s Fall Release 2025 delivers a consolidated set of upgrades across measurement accuracy, cross-platform LTV, attribution logic, AI-driven workflows, creative operations, and data collaboration—bringing these functions into a more unified and operationally consistent marketing environment.
For teams that rely heavily on creative performance to drive outcomes, Segwise integrates with AppsFlyer’s MMP data to close the loop on what the release does not cover in depth: creative-level intelligence. With Segwise, UA and growth teams can pinpoint which creatives drive incremental results, detect fatigue early, benchmark performance across networks, and turn AppsFlyer’s attribution and LTV signals into actionable creative decisions.
If you want to analyze your creative performance with structured, granular insights, you can explore Segwise as a dedicated solution for creative-level analytics. Start a free trial.
FAQ's
1. Which LLMs work with AppsFlyer Fall Release 2025’s Model Context Protocol (MCP)?
MCP already connects to major LLMs such as ChatGPT, Claude, Cursor, and other MCP-compatible models so that you can query AppsFlyer data directly from your chosen LLM.
2. Will my customer data be exposed when using Signal Hub or MCP?
No, Signal Hub uses anonymized, aggregated, clean-room matching, and MCP returns answers without handing over raw personal data, keeping PII from being shared.
3. When can teams start using AppsFlyer Fall Release 2025 features, and how do they get access?
The Fall Release was announced on November 18, 2025, and many features are now available or in beta. Request a demo or join product betas from AppsFlyer’s product pages.
4. Does the Fall Release change attribution or fraud detection?
Yes, the release adds a real-time AI attribution layer that flags fraud patterns like click flooding and produces cleaner attribution decisions.
5. Do I have to replace my analytics or CDP to use AppsFlyer Fall Release 2025?
No, the platform is built to integrate with existing stacks via Data Locker, Push/Pull APIs, and clean-room activations, so it complements current analytics and CDP tools.
