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

# Introduction to Marketer

> Learn how a request moves from workspace context to a reviewed result or controlled action in Marketer.

Marketer gives your team one place to ask questions, follow marketing work, review evidence, and decide what moves forward. Start on **Home**, give the Marketing Agent a clear outcome, and follow the work until you reach a result or a decision.

## Start on Home

Open [Marketer](https://app.marketer.com/dashboard) and confirm the workspace name in the upper-left corner before you submit anything. The active workspace determines which connected accounts, Knowledge, campaigns, briefs, creatives, products, and controls the agent can use.

<Frame caption="The Home canvas combines performance context, the request composer, active work, and recent summaries.">
  <img src="https://mintcdn.com/marketer-com/Zy37cwRIaZeFRNpC/assets/product-flows/home-overview.png?fit=max&auto=format&n=Zy37cwRIaZeFRNpC&q=85&s=4de37cc5d8346f9734837363b5598b3c" alt="Marketer Home canvas with performance overview, request composer, current work, daily summaries, plans, and reports" width="1020" height="900" data-path="assets/product-flows/home-overview.png" />
</Frame>

Use Home in this order:

1. **Confirm the workspace.** Switch workspaces if the visible name is not the company you intend to work with.
2. **Check the available context.** The overview and recent work help you see whether you are starting from the expected business state.
3. **Write the outcome.** Use **Ask anything or @ to add context** to describe the question, artifact, or decision you need.
4. **Attach the exact record when scope matters.** Select **+** and choose a campaign, brief, creative, or product instead of relying on a similar name. Use `@` to mention an agent.
5. **Follow the work.** A request can answer in chat, continue as a task, ask for clarification, or stop on a proposal that needs your decision.

## The product flow

Every useful request moves through the same five layers, even when the work itself differs.

| Layer         | What you provide or inspect                                                | What should be clear before you continue                                     |
| ------------- | -------------------------------------------------------------------------- | ---------------------------------------------------------------------------- |
| **Workspace** | The company, members, connected accounts, Knowledge, and settings in scope | You are in the intended workspace and the required source is connected.      |
| **Request**   | Outcome, source, scope, date range, constraints, and action boundary       | Marketer can tell what to produce and what it must not change.               |
| **Work**      | A direct response, task steps, specialist work, or a proposal              | You can see whether work is running, waiting for input, or ready for review. |
| **Result**    | The answer, report, brief, creative, workflow, or campaign work produced   | The result matches the request and identifies the evidence that supports it. |
| **Decision**  | Continue, revise, dismiss, approve, or verify in the owning surface        | A recommendation is separate from permission to change an external account.  |

### Answer flow

Use this for analysis and other read-only work:

1. Ask a bounded question on **Home**.
2. Let Marketer read the connected source and workspace context.
3. Open the completed result.
4. Verify its source, dates, scope, calculation, and limitations.
5. Ask a follow-up or continue in the relevant product area.

### Action flow

Work that can create or change something has an additional decision boundary. Marketer can show a proposed sequence before it runs.

At this state:

* Choose the displayed **Cancel** control when the proposed work should not continue.
* Choose **Edit** or **Edit brief** when the goal or steps need correction.
* Choose the displayed run control—such as **Accept & Run**, **Approve & Run**, or **Run it**—only after the scope, source, output, and downstream effect are correct.

The available controls depend on the task, your role, connected provider, guardrails, and workspace autonomy settings. A proposal is not evidence that anything has already changed.

## Know where work belongs

| You want to…                                        | Start or continue in…                                           | What you verify there                                 |
| --------------------------------------------------- | --------------------------------------------------------------- | ----------------------------------------------------- |
| Ask a question or brief the agent                   | **Home**                                                        | Request, attachments, and response                    |
| Reopen a conversation or longer task                | **Chats and tasks**, or the task link shown in the conversation | Current step, input request, and final result         |
| Inspect campaign structure or prepare campaign work | **Campaigns**                                                   | Provider account, campaign object, state, and metrics |
| Explore performance                                 | **Analytics**                                                   | Date window, source, filters, and metric definitions  |
| Create or review creative work                      | **Briefs**, **Creatives**, and **Studio**                       | Brand context, output, and intended use               |
| Make recurring work repeatable                      | **Workflow**                                                    | Trigger, inputs, destination, and review boundary     |
| Connect data                                        | **Integrations**                                                | Authorization, selected resources, sync, and coverage |
| Resolve a pending decision                          | **Awaiting action**                                             | Exact proposal, evidence, guardrail, and approver     |

Names and available controls can vary with your workspace role, plan, connected accounts, and product state. Use the surface shown in your workspace as the source of truth.

## Your first complete loop

For the safest first session:

1. [Set up your workspace and data](/get-started/workspace-and-data).
2. [Build a plan-only launch task](/get-started/first-task) from approved evidence and one connected ads account.
3. [Review the result](/get-started/review-result) before you use it.
4. Move into campaign, creative, analytics, or workflow work only after the first source is trustworthy.

<CardGroup cols={2}>
  <Card title="Complete the quickstart" icon="rocket" href="/get-started/quickstart">
    Follow the shortest path from sign-in to a verified result.
  </Card>

  <Card title="Learn the request model" icon="message-square" href="/essential-guidelines/effective-requests">
    Give the agent the outcome, evidence boundary, and authority it needs.
  </Card>
</CardGroup>


## Related topics

- [Working with Marketer](/working-with-marketer/overview.md)
- [Slack](/integrations/slack.md)
- [Review results](/essential-guidelines/review-results.md)
- [Navigation and Quick actions](/working-with-marketer/navigation-and-quick-actions.md)
- [History and notifications](/working-with-marketer/history-and-notifications.md)
