> ## Documentation Index
> Fetch the complete documentation index at: https://agenticadvertisingorg-feature-feedback.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

Signal providers have data. AI agents need access. The Signals Activation Protocol connects them.

This protocol enables AI assistants to discover, activate, and manage data signals through natural language—making it practical to navigate and utilize the vast universe of available segments.

## The Challenge: Navigating Hundreds of Thousands of Segments

Modern data platforms offer incredible depth—often hundreds of thousands of pre-built segments. While this provides unprecedented targeting options, it also creates real challenges:

* **Limited visibility**: Provenance and methodology vary across providers
* **Inconsistent metrics**: Each platform reports quality and performance differently
* **Naming conventions**: Segment names don't always clearly convey their contents
* **Discovery friction**: Finding the right segment requires deep platform expertise
* **Fixed catalogs**: Pre-built segments may not precisely match campaign needs

## How AI Assistants Transform Signal Discovery

### Current Approach

```
Human → Navigate platform UIs → Select from available → Activate and test
```

* **Challenges**: Time-intensive, may miss better options, requires platform expertise

### With AdCP

```
Human describes need → AI evaluates all options → Providers can create custom → Streamlined activation
```

* **Benefits**: Comprehensive discovery, informed decisions, precise targeting

## The Power of AI-Assisted Discovery

### AI Can Process and Compare Metadata at Scale

```
Human: I need people likely to buy luxury cars in the next 30 days

AI: Analyzing 487 automotive segments across 3 providers...

Best match: DataVault's "Premium Auto Purchase Intent"
- Provenance: Dealer data + finance searches + config tools
- Freshness: Updated daily, 89% still in-market after 30 days
- Accuracy: 73% precision based on 6-month validation
- Size: 2.3M individuals (RampID, UID2, MAIDs)
- Performance: 4.2x lift vs. general auto intenders
- Price: $8 CPM

Alternatives with trade-offs:
- MotorSignals: Higher precision (81%) but smaller scale (890K)
- AutoData: Lower cost ($5 CPM) but monthly refresh only
```

### Providers Create Exactly What You Need

```
Human: I need business travelers who fly monthly but haven't 
        booked hotels through our platform

AI to DataVault: Creating custom segment from your requirements...

DataVault's AI: 
- Analyzing travel graph for monthly flyers
- Cross-referencing with your CRM (secure match)
- Excluding recent hotel bookers
- Building segment in real-time...

Created: "Your Brand - High-Frequency Flyers Without Hotel Bookings"
- Size: 147K matched individuals  
- Available on: RampID, UID2, MAIDs
- Exclusive to your account
- Refreshes daily from your CRM
```

This enables a shift from browsing pre-built catalogs to requesting exactly what you need.

## From Static Catalogs to Dynamic Capabilities

### Traditional Approach: Pre-Built Catalogs

* Hundreds of thousands of pre-defined segments
* Challenging to evaluate fit and quality across providers
* Long lead times for new segment creation
* Limited ability to customize for specific needs
* Decision-making based on available documentation

### Enhanced with AdCP: AI-Powered Discovery & Creation

```
Human: Find affluent parents interested in EVs but exclude Tesla owners

AI: Found 3 approaches:

1. Existing segment: FamilyData's "EV-Curious Affluent Parents"
   - Ready now, 340K individuals, $7 CPM
   - Note: Includes ~15% Tesla owners

2. Custom creation: DataVault can build exactly this
   - Combining: Affluent parents + EV interest - Tesla registration
   - 48-hour creation, 290K individuals, $9 CPM
   - 100% match to your criteria

3. Campaign combination: Use multiple signals
   - FamilyData's segment + negative targeting for Tesla
   - Immediate activation, handled at DSP level

Which approach works best for your campaign?
```

## Universal Signal Keys

AdCP supports every major identity and targeting methodology:

### Identity Signals

* **RampID**: LiveRamp's universal identifier
* **UID2**: The Trade Desk's open-source ID
* **Publisher IDs**: First-party authenticated users
* **Hashed emails**: SHA256 email matching

### Device Identifiers

* **MAIDs**: Mobile advertising IDs (IDFA/GAID)
* **Connected TV IDs**: Roku ID, Samsung ID, etc.
* **Cookie IDs**: Platform-specific cookies
* **IP addresses**: Household-level targeting

### Geographic Signals

* **Lat/Long**: Precise location polygons
* **Postal codes**: ZIP, ZIP+4, international formats
* **DMA**: Designated Market Areas
* **Custom geofences**: Store radii, event venues

### Multi-Dimensional Signals

* **Domain × Time**: "wsj.com visitors on weekday mornings"
* **Weather × Location**: "Cold fronts approaching major metros"
* **Content × Sentiment**: "Reading positive EV articles"
* **Behavior × Frequency**: "Daily coffee shop visitors"

### Temporal Signals

* **Real-time triggers**: "Temperature exceeds 85°F"
* **Time windows**: "Business hours in local timezone"
* **Recency**: "Visited site in last 24 hours"
* **Predictive**: "Likely to purchase in next 7 days"

## Key Use Cases

### 💡 Dynamic Weather-Based Activation

```
Human: Activate ice cream ads when tomorrow's weather will drive demand

AI: Creating dynamic weather signal:
- Monitoring 150 metro areas
- Trigger: Tomorrow's forecast > 85°F + sunny
- Currently qualifying: Miami, Phoenix, San Diego
- Will auto-activate/deactivate based on forecasts
- Refreshes every 6 hours

Signal is now live across all platforms.
```

### 🎯 Multi-Signal Campaign Orchestration

```
Human: I want to target crypto investors who also collect wine 
        and live in high-income neighborhoods

AI: Found complementary signals for your campaign:

FinanceSignals: "Active Crypto Traders"
- Available on: UID2, RampID, MAIDs
- Size: 1.8M individuals
- Price: $5 CPM

LuxuryData: "Wine Collectors" 
- Available on: RampID, hashed emails
- Size: 450K individuals
- Price: $6 CPM

CensusPlus: "HHI > $250K by Postal"
- Geographic signal: ZIP+4 level
- Coverage: All US metros
- Price: $4 CPM

Activating all three on DSP Platform X (using RampID).
Platform will handle audience intersection.
Total signal cost: $15 CPM combined
```

### 🔬 Real-Time Contextual Activation

```
Human: Target users reading about EVs right now on premium news sites

AI: Activating real-time contextual signal:
- Monitoring: WSJ, FT, Bloomberg, Reuters
- Content match: Electric vehicle articles
- Sentiment: Positive/neutral only
- Viewability: > 70%
- Updates: Every 5 minutes

Current inventory: 2.3M impressions/day
Activating on all connected SSPs...
```

### 📊 Custom Signal Creation from Your Data

```
Human: Can DataVault create a custom signal from our CRM data 
        for high-value customers who haven't purchased recently?

AI: DataVault can create a custom signal using your data:

"Your Brand - Lapsed High-Value Customers"
- Input: Your CRM list (hashed emails)
- Match against DataVault's identity graph
- Output formats: RampID, UID2, MAIDs
- Expected match rate: 60-70%

Signal characteristics:
- Size: ~127K matched individuals
- Refresh: Daily from your CRM
- Exclusive to your account
- Price: $8 CPM

DataVault will score recency and value, making it
available across all your connected platforms.

Proceed with signal creation?
```

## Multi-Platform Intelligence

See your entire signal ecosystem in one view:

```
Human: Show me all my travel signals across platforms

AI: Your travel signals deployment status:

**"Luxury Travel Intenders" (450K users)**
✅ The Trade Desk: ttd_lux_travel_q4 (RampID)
✅ Amazon DSP: amzn_travel_lux_2024 (UID2)
✅ Google DV360: dbm_luxury_travel (Cookie ID)
⏳ Yahoo DSP: Activating... (2 hours)

**"Last-Minute Bookers" (1.2M users)**
✅ All platforms active
Using: RampID, UID2, and MAIDs

**"Business Travelers" (890K users)**
Custom segment you created last week
Now used by 47 campaigns
Generating $12K/day in signal revenue
```

## Core Tasks

The Signals Activation Protocol supports two primary tasks:

### 1. [get\_signals](/dist/docs/2.5.3/signals/tasks/get_signals)

Discover signals across providers using natural language, with visibility into available identity keys.

### 2. [activate\_signal](/dist/docs/2.5.3/signals/tasks/activate_signal)

Deploy signals to platforms with automatic identity resolution.

## How It Works

```mermaid theme={null}
flowchart LR
    A[Natural Language Request] --> B[AI Understands Intent]
    B --> C[Search Across Providers]
    C --> D{Signal Exists?}
    D -->|No| E[Provider Creates Custom]
    D -->|Yes| F[Analyze Quality & Metadata]
    E --> G[Match Identity Keys]
    F --> G
    G --> H[Activate on Platforms]
    H --> I[Campaign Uses Signals]
    I --> J[Monitor & Optimize]
```

## The Value Proposition

**For Signal Providers:**

* 📈 **Increased utilization**: Better discovery leads to more signal activation
* ⚡ **Faster time-to-market**: Create custom signals more efficiently
* 🎯 **Better matching**: Build signals that precisely match buyer needs
* 🌍 **Broader compatibility**: Support multiple identity types and platforms

**For Advertisers:**

* 🔍 **Better visibility**: Access to comprehensive metadata and performance data
* 🎨 **Precise targeting**: Request and receive exactly what you need
* 🤖 **Informed decisions**: AI helps evaluate and compare options
* 🚀 **Faster activation**: Streamlined discovery and deployment process

## Next Steps

* 📖 **Technical Teams**: Review the [Protocol Specification](/dist/docs/2.5.3/signals/specification)
* 💻 **Developers**: Explore the [Reference Implementation](https://github.com/adcontextprotocol/signals-agent)
* 🏗️ **Platform Providers**: See [Integration Guide](/dist/docs/2.5.3/signals/tasks/get_signals)
* 💬 **Everyone**: Join the [Community](https://join.slack.com/t/agenticads/shared_invite/zt-3c5sxvdjk-x0rVmLB3OFHVUp~WutVWZg)

***

*The Signals Activation Protocol is part of the broader [AdCP ecosystem](/dist/docs/2.5.3/intro#the-adcp-ecosystem-layers), transforming how advertising technology works together.*