> ## Documentation Index
> Fetch the complete documentation index at: https://developers.criteo.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Example Workflows

> Common agent workflows against the Criteo Retail Media MCP server.

## Introduction to Workflows

These workflows illustrate how MCP can be used to explore data, analyze performance, and support decision-making across different user roles.

They are intended as **starting points for testing and validation**, not as an exhaustive list of capabilities. You can use the suggested prompts directly in your MCP-compatible client and iterate with follow-up questions to refine results.

Start with the shared validation workflow to confirm access and scope, then move into more advanced workflows based on your use case.

<Info>
  Most workflows benefit from follow-up prompts to refine results.
</Info>

<Columns cols={3}>
  <Card title="Workflow 1" href="/mcp/docs/retail-media/example-workflows#workflow-1-shared-account-and-setup-validation">
    Shared Account and Setup Validation
  </Card>

  <Card title="Workflow 2" href="/mcp/docs/retail-media/example-workflows#workflow-2-global-capability-and-retailer-discovery">
    Global Capability and Retailer Discovery
  </Card>

  <Card title="Workflow 3" href="/mcp/docs/retail-media/example-workflows#workflow-3-global-performance-and-competitiveness-monitoring">
    Global Performance and Competitiveness Monitoring
  </Card>

  <Card title="Workflow 4" href="/mcp/docs/retail-media/example-workflows#workflow-4-brand-campaign-performance-pull">
    Brand Campaign Performance Pull
  </Card>

  <Card title="Workflow 5" href="/mcp/docs/retail-media/example-workflows#workflow-5-brand-line-item-diagnostics">
    Brand Line-item Diagnostics
  </Card>
</Columns>

***

## Workflow 1: Shared Account and Setup Validation

**Persona**: Global Commerce and Brand Users

**Goal**: Validate that auth, scoping, and discovery work correctly before deeper testing.

**Suggested prompts**

* "Which accounts are visible through this MCP connection?"
* "Are the expected retailers, brands, and creatives accessible?"
* "Is the retrieved scope aligned with the user’s assigned brands / accounts?"

**What good looks like**

* Confirms the right data scope for both personas
* Surfaces access issues early

***

## Workflow 2: Global Capability and Retailer Discovery

**Persona**: Global Commerce

**Goal**: Validate that global users can understand account scope, retailer availability, and supported capabilities across accounts / retailers.

**Suggested prompts**

* "Which retailers are available across the accounts I can access?"
* "Which retailers support Onsite Display and what formats are available?"
* "Show me retailer capability differences across the accounts in scope."

**What good looks like**

* Useful for portfolio-level planning and partner enablement
* Output is structured and easy to compare across retailers

***

## Workflow 3: Global Performance and Competitiveness Monitoring

**Persona**: Global Commerce

**Goal**: Validate whether global users can get broad performance and share-of-voice visibility without manual exports.

**Suggested prompts**

* "Summarize account-level performance over the last 30 days."
* "Which retailers or categories show weaker share of voice?"
* "Where are we seeing the biggest performance changes across the portfolio?"

**What good looks like**

* Useful for global / portfolio-level monitoring
* Helps identify where deeper brand-level investigation is needed

***

## Workflow 4: Brand Campaign Performance Pull

**Persona**: Brand Users

**Goal**: Validate brand-scoped campaign reporting for day-to-day optimization review.

**Suggested prompts**

* "Summarize campaign performance for Brand X over the last 14 days."
* "Which campaigns had the highest ROAS last month for this brand?"
* "Which campaigns are underperforming vs the previous period?"
* “Which campaigns are underpacing and how can I increase their pacing?”

**What good looks like**

* Brand users can quickly understand what changed
* Output is usable without manual data cleanup

***

## Workflow 5: Brand Line-item Diagnostics

**Persona**: Brand Users

**Goal**: Validate that line-item reporting is useful for troubleshooting and optimization support.

**Suggested prompts**

* "Which line items are dragging performance down for Brand X?"
* "Show me underdelivering line items from the last 7 days."
* "Which line items have high spend but weak ROAS?"

**What good looks like**

* Clear, actionable diagnosis at brand / line-item level
* Users can identify what needs attention without stitching multiple exports together

***

<br />