> ## Documentation Index
> Fetch the complete documentation index at: https://agenticadvertisingorg-snap-format-preview-links.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# S3: Signals and audiences
> AdCP specialist module S3: Signals and audiences. Signal discovery, activation, privacy controls, and optimization loops with sandbox data providers across six industries.
# S3: Signals and audiences
**Members only** — Requires Practitioner credential. \~45 minutes with Addie. Combines hands-on lab and adaptive exam.
This specialist module tests your mastery of the signals protocol. You'll work with sandbox signal providers — automotive data, geo/mobility, retail purchase data, identity/demographics, publisher contextual signals, and CDP audiences — to discover signals, activate them, manage privacy, and design optimization loops. Addie adapts the experience to your role in the ecosystem.
Passing earns the **AdCP specialist — Signals** credential.
## Specialisms this track prepares you to validate
The following `specialisms` fall under the `signals` domain. Each has its own compliance storyboard — see the [Compliance Catalog](/docs/building/compliance-catalog) for the full taxonomy.
| Specialism | Status | What it covers |
| -------------------- | ------ | --------------------------------------------------- |
| `signal-owned` | stable | Owned signal agent exposing first-party segments |
| `signal-marketplace` | stable | Marketplace signal agent reselling third-party data |
**Scope of AdCP signals**
Before investing time in training, it helps to know what the protocol covers today and where boundaries exist.
* **In scope**: Identity-derived attributes (income tiers, life stages), behavioral signals (purchase intent, visit frequency), contextual signals (content category, sentiment), geographic audiences (trade areas, store visitors)
* **Not yet in scope**: Identity resolution and matching (linking devices to people), real-time geofencing triggers (push when someone enters a zone), measurement and attribution pipelines
If your company works in the "not yet" areas, you can still publish the **audience segments** your data produces — the protocol covers the targeting output even when it doesn't cover the underlying infrastructure.
## What you'll demonstrate
* Discover and evaluate signals from multiple provider types (data providers, retailers, publishers, CDPs)
* Activate appropriate signals for different campaign objectives
* Understand signal value types (binary, categorical, numeric) and how they affect targeting
* Manage audience activation including privacy considerations and deactivation
* Configure event tracking with `sync_event_sources` and `log_event`
* Read `pricing_options[]` from `get_signals` responses, pass the selected `pricing_option_id` through `activate_signal` and `report_usage`
* Design optimization loops using signals and delivery data
* Reason about the signals ecosystem — who provides data, who consumes it, and how authorization works
**Vendor pricing is consistent across protocols** — All vendor services use `pricing_options[]` on discovery responses and `pricing_option_id` in `report_usage`. Vendors may offer multiple options — volume tiers, context-specific rates, or different models per product line. Signals, creative, content standards, and property list agents all follow the same pattern. See [S2: Creative mastery](/docs/learning/specialist/creative) for the creative pricing model.
## Prerequisite reading
### Core signals tasks
Signal discovery: find targetable audiences, contextual categories, and measurement data.
Activate signals for campaign targeting or measurement.
### Supporting concepts
The signals protocol: audience segments, contextual signals, measurement, and optimization.
Formal specification for the signals protocol.
How data providers publish signal definitions via adagents.json.
How different company types — retailers, publishers, CDPs, identity companies — participate in signals.
Event sources, `log_event`, and attribution setup.
How signals feed into campaign optimization decisions.
## Connecting to the test agent
Lab exercises run against the public test agent. Use the shared token — no signup required:
```bash theme={null}
export ADCP_AUTH_TOKEN="1v8tAhASaUYYp4odoQ1PnMpdqNaMiTrCRqYo9OJp6IQ"
export AGENT_URL="https://test-agent.adcontextprotocol.org/signals/mcp"
```
See the [Quickstart](/docs/quickstart) for a walkthrough of your first call.
## Lab exercises
The sandbox training agent includes signal providers representing different ecosystem roles. You'll work with all of them:
| Provider | Type | Signals |
| ------------------------- | ------------- | -------------------------------------------------------------------------------------------------- |
| Trident Auto Data | Data provider | EV buyers, vehicle ownership, purchase propensity, service due, service history |
| Meridian Geo | Geo/mobility | Competitor visitors, visit frequency, trade area, commute pattern, dwell time, day-part visitation |
| ShopGrid Shopper Insights | Retailer | Category buyer, loyalty tier, basket value, new to brand, purchase frequency, brand affinity |
| Keystone Identity | Identity | Household income, life stage, cross-device reach, credit activity, household composition |
| Pinnacle News Signals | Publisher | Content category, engaged reader, subscriber tenure, sentiment, page type |
| Prism CDP | CDP | High LTV, cart abandoner, engagement score, churn risk, cross-device |
### Exercise 1: Signal discovery
Query the sandbox signals agent for signals matching different campaign objectives. Compare signals across provider types — what does a data provider's automotive signal look like versus a retailer's purchase signal?
### Exercise 2: Signal activation
Activate signals for a sandbox campaign. Observe how activation keys work and how deployment status changes.
### Exercise 3: Audience management
Activate and deactivate signals. Consider privacy: when does a signal need to be deactivated? How does consent affect signal availability?
### Exercise 4: Ecosystem scenarios
Addie will present a scenario from a specific perspective — you might be building published signal definitions for a retail media network, choosing signals for an agency's client campaign, or designing a CDP integration. Apply your protocol knowledge to the scenario.
### Exercise 5: Build published signal definitions (provider perspective)
The previous exercises focus on the buyer side — discovering and activating signals. This exercise shifts to the provider side. Construct an `adagents.json` signals entry for a fictional data provider of your choice (geo, retail, identity, etc.). Your definitions should include:
* At least one signal of each value type (`binary`, `categorical`, `numeric`)
* Descriptive IDs, tags, and metadata following the [data provider guide](/docs/signals/data-providers)
* An `authorized_agents` entry authorizing a signals agent to resell your signals
* `restricted_attributes` declarations on signals derived from sensitive personal data (e.g., `health_data`, `racial_ethnic_origin`)
* `policy_categories` declarations on signals that carry regulatory implications (e.g., `children_directed`, `fair_housing`)
Verify your definitions by checking that the signals appear correctly in `get_signals` results when queried by tag or description, and that governance attributes are preserved in the signal metadata.
## Assessment
| Dimension | Weight | What Addie evaluates |
| ----------------------- | ------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Protocol mastery | 25% | Complete signals lifecycle (discovery → activation → targeting → deactivation) |
| Privacy compliance | 20% | Handles consent, deactivation, and data governance correctly |
| Measurement skill | 20% | Configures conversion tracking and attribution |
| Ecosystem understanding | 20% | Explains how different provider types fit into signals |
| Ecosystem scenarios | 15% | Constructs valid published signal definitions, understands both buyer and provider perspectives, reasons about activation destinations (agent vs platform) |
Passing threshold: 70%.
## Start this module
"I'd like to start the signals specialist module."