# AppsFlyer (Pull API) MCP for AI Agents MCP

> AppsFlyer (Pull API) provides direct access to your mobile attribution and marketing analytics data. Use this MCP to pull raw reports on installs, in-app events, uninstalls, and aggregate performance metrics based on geography or media source. You can ask for specific campaign data by defining precise date ranges, helping you track user behavior and calculate lifetime value with natural language prompts.

## Overview
- **Category:** marketing-automation
- **Price:** Free
- **Tags:** mobile-attribution, marketing-analytics, event-tracking, performance-reporting, user-acquisition-data, data-retrieval

## Description

Your AI client connects directly to your mobile attribution data through this MCP, giving it a comprehensive view of your marketing performance. Instead of jumping between dashboards or exporting files manually, you talk to your agent and ask for the metrics you need—whether it's daily totals for an entire month or raw event logs from a specific partner source.

For instance, if you need to audit how well 'Facebook Ads' performed last quarter against 'Google Search,' your agent pulls those numbers instantly. You can get granular reports on non-organic installs and track every single in-app action (like purchases or level completions) by date. The data comes back clean, structured as CSV files ready for analysis. This means you spend less time gathering data and more time making decisions. Because Vinkius hosts this MCP, your agent gets secure access to all of your attribution data through one connection point.

## Tools

### get_account_check
Checks the current connection status to ensure your AppsFlyer account is linked correctly.

### get_daily_report
Generates a summary report showing how well campaigns performed on a specific day.

### get_geo_report
Provides an aggregate view of performance metrics, broken down by geographical location.

### get_in_app_events_report
Downloads a detailed report containing raw records for user actions inside your app, like purchases or level completions.

### get_installs_report
Retrieves the complete dataset detailing all non-organic app installs that occurred during the specified time frame.

### get_partners_report
Compiles aggregate performance summaries, allowing you to compare different media source partners directly.

### get_uninstalls_report
Retrieves the raw data report detailing how many users uninstalled the app over time.

## Prompt Examples

**Prompt:** 
```
What were the top 3 media sources for installs last month?
```

**Response:** 
```
**AppsFlyer Performance Summary: Last 30 Days**

| Partner Source | Installs (Non-Org) | Sessions | Conversion Rate |
| :--- | :--- | :--- | :--- |
| Google Search | 12,500 | 45,000 | 4.1% |
| Facebook Ads | 8,900 | 32,100 | 3.5% |
| TikTok | 4,100 | 15,000 | 2.8% |

*Source: `get_partners_report`*
```

**Prompt:** 
```
Did our purchases drop in the Northeast region last week?
```

**Response:** 
```
**Geographic Event Analysis (Last 7 Days)**

The Northeast region saw a dip in purchase events. Here's the breakdown:
*   **Total Purchases:** 450 units (down from 612).
*   **Top Events:** 'Premium Subscription' and 'Ad Removal.'
*   **Key Takeaway:** The drop correlates with low activity from the 'University' metro area. Check campaigns targeting that zone.
```

**Prompt:** 
```
Give me all raw data for level complete events today.
```

**Response:** 
```
**Raw Event Log: Level Complete (Today)**

```json
{
  "event": "level_complete",
  "user_id": "xyz789",
  "timestamp": "2026-06-12T14:35:00Z",
  "details": {"level": 12, "difficulty": "hard"}
}
```

This data is ready for export and contains all necessary user details.
```

## Capabilities

### Verify account connection
Confirms that the AI client can securely connect to and authenticate with the AppsFlyer platform.

### Generate daily performance summaries
Pulls a high-level report showing aggregated metrics for any specific day you request.

### Analyze geographic trends
Generates aggregate reports detailing how campaign performance varies by user location or geography.

### Extract raw event data
Downloads detailed, non-aggregated records for specific in-app events like purchases or level completions.

### Retrieve install reports
Gets a full dataset of all non-organic installs and uninstalls over a specified period.

### Track media source performance
Compiles aggregate metrics to compare how different advertising partners or sources performed against each other.

## Use Cases

### Diagnosing a drop in sales after a campaign launch
A Growth Marketer notices sales dipped last week. They ask their agent to pull the `get_in_app_events_report` for that specific week, filtering by 'purchase' events. The report immediately shows the dip correlated with a change in media source spending, pointing the team toward the problem area.

### Comparing two rival ad campaigns side-by-side
The UA Manager needs to know if Campaign A or Campaign B generated more quality installs. They use `get_partners_report` and ask for a comparison of 'installs' metrics between the two media sources, providing clear numbers that justify budget reallocation.

### Investigating user drop-off points
A Data Analyst suspects users are leaving after downloading. They use `get_uninstalls_report` combined with an installs report to quantify the churn rate and pinpoint if certain geographic areas show unusually high uninstallation rates.

### Creating a monthly performance summary for leadership
A Marketing Director needs a simple, executive-level view. They prompt their agent for `get_daily_report` over the last 30 days and ask it to summarize key metrics, generating a single clean, easy-to-read markdown output.

## Benefits

- Get full visibility into campaign performance. Use the `get_partners_report` to directly compare installs across different media sources, eliminating manual spreadsheet comparisons.
- Deep dive into user behavior. The `get_in_app_events_report` pulls raw data for purchases or level completions, letting you find specific patterns that surface in standard dashboards.
- Save time on reporting. Instead of generating reports day by day, use the `get_daily_report` to pull entire monthly summaries with a single prompt.
- Track user churn and growth. The combination of `get_installs_report` and `get_uninstalls_report` lets you calculate net retained users instantly.
- Analyze market reach. Need to know if your campaign focused on Germany or Brazil performed better? Use the `get_geo_report` for instant, location-based insights.

## How It Works

The bottom line is that your agent translates complex marketing questions into actionable data pulls from AppsFlyer, bypassing manual dashboard work entirely.

1. You tell your AI client exactly what data you need, for example, 'Show me the performance summary for EMEA last month.'
2. The agent uses this MCP to send a secure request to AppsFlyer's Pull API, specifying the desired report type (e.g., geographic) and date range.
3. Your AI client receives the requested data set, formatted as structured CSV files that you can immediately read or process.

## Frequently Asked Questions

**How can I use the AppsFlyer (Pull API) MCP to compare different ad channels?**
You can pull a direct comparison of metrics from various advertising sources. By using the agent, you get an aggregate report that shows installs and sessions side-by-side for every partner, helping you instantly see where your budget should shift.

**Is this MCP useful if I only care about purchases and events?**
Absolutely. You don't need to sift through everything. You can specifically ask the agent to retrieve a raw data report for 'purchase' or 'level_complete' events, giving you exactly the transactional details you want for deeper analysis.

**Can I use this MCP to analyze user retention and churn?**
Yes. You can combine the `get_installs_report` with the `get_uninstalls_report`. This lets your agent calculate net retained users over a period, which is key for figuring out if your marketing spend is actually keeping people engaged.

**Does AppsFlyer (Pull API) MCP handle data from different countries?**
It does. You can request reports filtered by geography using the `get_geo_report`. This lets you compare performance metrics between regions, helping you understand which markets are most profitable.

**What if I need historical data for a very specific date range?**
You simply tell your agent the exact start and end dates. The system handles the date parameters automatically when pulling reports like the daily performance summary, ensuring you focus only on the period you care about.