# Meta Ads MCP

> Meta Ads MCP gives your AI agent direct access to Facebook and Instagram advertising data. Manage campaigns, track performance metrics like ROAS, and build custom audiences—all without manually opening Meta Ads Manager. Your agent can list all running campaigns, analyze ad set spending, or compare creative performance just by answering a natural language prompt.

## Overview
- **Category:** industry-titans
- **Price:** Free
- **Tags:** ad-campaign-management, performance-analytics, social-advertising, audience-targeting, roas-tracking, ad-spend-optimization

## Description

Managing ads across Facebook and Instagram usually means bouncing between dashboards, exporting messy spreadsheets, and juggling multiple tabs to stitch together one coherent story. This MCP changes that. You connect Meta Ads to your agent and manage everything conversationally. Your AI client reads the ad account details, pulls campaign insights, and analyzes every metric you care about—from total spend to conversion rate. Need to know which audience is underperforming? Ask. Want to pause a losing effort or activate a winner? Tell it. Everything runs through one unified data source. By integrating Meta Ads via Vinkius, your agent can handle complex tasks like generating lookalike audiences and comparing the performance of ad creatives across all campaigns. You get actionable insights instantly, making optimization a simple conversation.

## Tools

### list_campaigns
Retrieves a list of all active and paused advertising campaigns in the account.

### get_account_info
Pulls basic details about the overall ad account, like name and status.

### get_campaign_insights
Fetches detailed performance metrics for a specific campaign, including reach, clicks, and cost per result.

### list_ad_sets
Lists all ad sets that belong to a particular advertising campaign.

### list_ads
Retrieves a list of specific ads within an ad set for review.

### get_account_insights
Gathers high-level performance data across the entire account, useful for quarterly reviews.

### list_audiences
Lists existing custom audiences that you can use to target ads.

### list_creatives
Retrieves a list of ad creatives, allowing you to compare image or video performance.

### pause_campaign
Stops an advertising campaign immediately; it can be reactivated later.

### activate_campaign
Restarts a previously paused campaign, resuming its ad spend and visibility.

## Prompt Examples

**Prompt:** 
```
What's my Meta ad spend and ROAS this month?
```

**Response:** 
```
This month: Total spend: $15,400. Reach: 2.1M. Impressions: 5.8M. Clicks: 42,000. CTR: 0.72%. CPC: $0.37. Purchases: 620. ROAS: 4.8x. Best campaign: 'Product Catalog Retargeting' (8.2x ROAS).
```

**Prompt:** 
```
Which ad creatives are performing best?
```

**Response:** 
```
Top 3 creatives by ROAS: 1) Video 'Product Unboxing' (6.5x ROAS, $2.1K spend), 2) Carousel 'Summer Collection' (5.8x, $3.4K), 3) Static 'Testimonial - Sarah' (5.2x, $1.8K). Worst: 'Brand Story v3' (0.9x ROAS — consider pausing).
```

**Prompt:** 
```
Create a lookalike audience from my top 1000 customers.
```

**Response:** 
```
Lookalike audience created: 'Top 1K Customers - 1% Lookalike' in US. Estimated reach: 2.5M people. Similarity: 1% (highest match). Ready to use in campaigns.
```

## Capabilities

### Analyze campaign health
The agent lists every active campaign and retrieves detailed account-level spending and performance data.

### Build specific audiences
You generate custom or lookalike audiences based on existing customer lists, ready for immediate ad targeting.

### Optimize campaign status
The agent can pause an underperforming campaign or reactivate a previously suspended one with simple instructions.

### Compare creative performance
You list and compare ad creatives, identifying which images or videos generate the highest return on ad spend (ROAS).

## Use Cases

### Stopping the bleed on low-performing ads
A brand owner notices their spend is high but conversions are low. They ask their agent, 'Show me all campaigns with a CPA over $10.' The agent uses `get_campaign_insights` and identifies three struggling campaigns, allowing the user to pause them immediately using `pause_campaign`, saving budget.

### Launching a new product line campaign
An agency needs to target existing high-value customers. They instruct their agent to 'Create a lookalike audience based on our top 5,000 buyers.' The agent uses `list_audiences` and generates the necessary custom audience for the new product launch.

### Quarterly budget review
A marketing director needs an overview of account health. They ask their agent to summarize performance, triggering `get_account_insights`. The system returns total spend, reach, and ROAS across all time periods for easy presentation.

### Debugging a campaign failure
A marketer suspects an ad set is failing because the creative isn't resonating. They use `list_ad_sets` to locate the area, then ask the agent to compare the performance of three different ads using `get_campaign_insights`, pinpointing the weakest creative.

## Benefits

- You can analyze campaign performance by requesting `get_campaign_insights` or pulling high-level data with `get_account_insights`. This eliminates the need to open multiple reports and manually calculate ROAS.
- Building targeted traffic is fast. You use `list_audiences` to see what custom audiences you have, then ask your agent to create a new lookalike audience based on top customers.
- When performance dips, you don't waste time clicking through the ad hierarchy. You tell your agent to check if a campaign needs adjusting and can instantly pause it using `pause_campaign` or reactivate it with `activate_campaign`.
- Comparing what works takes seconds. Instead of downloading CSVs for every piece of media, you ask the MCP to list creatives (`list_creatives`) and compare their performance side-by-side.
- The agent keeps track of everything. You can get a comprehensive view by listing all campaigns using `list_campaigns`, then drill down into specific ad sets with `list_ad_sets` for granular review.

## How It Works

The bottom line is that your AI client treats the entire ad network like an internal database, letting you query it naturally instead of clicking through menus.

1. First, subscribe to the Meta Ads MCP via Vinkius and create an app in your developer account.
2. Next, you simply ask your AI client what you need—for example, 'Show me campaigns with ROAS below 3x.'
3. The agent uses this MCP to fetch performance metrics and campaign details directly from Meta's system, giving you a clear answer.

## Frequently Asked Questions

**How do I get my Meta Ads API credentials?**
1. Go to **developers.facebook.com** and log in.
2. Click **My Apps** → **Create App**.
3. Select **Business** as the app type.
4. Fill in the app name and connect it to your Business Manager.
5. Add the **Marketing API** product to your app.
6. Go to **Settings → Basic** and copy your **App ID** and **App Secret**.
7. Generate a **System User Access Token** in Business Manager (Business Settings → System Users → Generate Token).
8. Paste all values below.

💡 For production use, submit your app for **App Review** to get stable, long-lived tokens.

**Can I manage both Facebook and Instagram ads?**
Yes. Meta Ads Manager unifies Facebook and Instagram advertising. Your AI agent manages campaigns across both platforms through the same API — including placements in Feed, Stories, Reels, and Messenger.

**Can I create lookalike audiences?**
Yes. Your AI agent can create lookalike audiences from customer lists, website visitors, or app users — with configurable similarity percentages (1% being most similar, 10% broadest reach).

**What conversion metrics can I track?**
Track purchases, add-to-cart, initiate checkout, lead submissions, app installs, registrations, page views, and custom conversion events — with attribution windows of 1-day click, 7-day click, and 1-day view.