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

# Actions

> Ranked recommendations to improve your AI shopping visibility.

<Info>
  Actions are available on paid plans.
</Info>

The **Actions** page turns diagnostics into an execution queue. It is where analysis becomes work.

<Frame caption="Actions queue showing severity-ranked recommendations with state badges">
  <img src="https://mintcdn.com/autonome/WzX-ugIsBIOg_KpO/images/actions.jpg?fit=max&auto=format&n=WzX-ugIsBIOg_KpO&q=85&s=01e22df73015a3d2f1b06b067c1adb64" alt="Actions page with a summary bar at the top (Active, Quick wins, Completed, Verified counts) and a list of action cards below, each showing title, severity badge, difficulty, and affected prompts" width="1475" height="812" data-path="images/actions.jpg" />
</Frame>

For catalog-wide readiness work, Actions summarizes the recurring issue and links directly into the matching view on the [Products](/features/products) page.

## What you see first

At the top of the page, Actions summarizes the current queue:

* **Active**
* **Quick wins**
* **Completed**
* **Verified**

This gives you a fast sense of how much work is open and how much has already been addressed.

## Reading an action card

Each action includes:

| Field                | Meaning                                                                                       |
| -------------------- | --------------------------------------------------------------------------------------------- |
| **Title**            | What to do, in plain terms                                                                    |
| **Severity**         | Deterministic priority based on prompt coverage plus fixed floors for important issue buckets |
| **Difficulty**       | How much effort is required to fix it                                                         |
| **Execution**        | Whether it is automatable or manual                                                           |
| **State**            | Whether it is open, in progress, completed, verified, or dismissed                            |
| **Affected prompts** | Which tracked queries this fix will impact                                                    |
| **Expected impact**  | Why this action matters in practice                                                           |

## Action categories

| Category       | What it involves                                                |
| -------------- | --------------------------------------------------------------- |
| **Catalog**    | Add or update product metafields, attributes, tags, or taxonomy |
| **Content**    | Rewrite product titles, descriptions, or collection page copy   |
| **Schema**     | Add or fix JSON-LD structured data on product pages             |
| **Outreach**   | Pitch products for editorial coverage or third-party reviews    |
| **Collection** | Create or restructure collections to match how buyers search    |

## Filters and sorting

You can filter the queue by:

* **State**
* **Severity**
* **Type**
* **Difficulty**
* **Execution**

You can also sort by **highest impact**, **newest**, or **easiest**.

This makes it easier to split the queue by owner, urgency, or execution style.

## Quick wins and backlog

Actions that are both high-leverage and relatively easy are called **Quick wins**. They appear separately from the broader backlog so teams can move fast on the highest-return work.

Everything else stays in the main backlog for ongoing prioritization.

## Severity levels

| Severity     | Meaning                                                      |
| ------------ | ------------------------------------------------------------ |
| **Critical** | Directly blocking you from appearing in high-value prompts   |
| **High**     | Significantly reducing your rank in prompts where you appear |
| **Medium**   | Improvement opportunity with measurable impact               |
| **Low**      | Minor polish — worth doing, but not urgent                   |

## How to use Actions

**Start with Quick wins and High severity.** These usually represent the best tradeoff between impact and effort.

**Then open the affected products.** Catalog-wide readiness actions include a **View affected products** shortcut so you can inspect the exact product rows, issue evidence, and Shopify or storefront links before making changes on the [Products](/features/products) page.

**For catalog and content actions, use Enrichment.** The [Enrichment](/features/enrichment) workflow can generate AI-suggested rewrites, structured tags, SEO fields, metafields, and alt text for the affected products — so you review proposals rather than writing every fix by hand.

**Then work by owner.** Catalog and schema fixes often belong to product or engineering, while content and outreach work may belong to marketing or SEO.

**Use state transitions actively.** Move actions through open, in progress, completed, verified, or dismissed so the queue stays trustworthy.

<Tip>
  After implementing a batch of actions, use [Overview](/features/overview) to watch metric movement, [Analysis](/features/analysis) to confirm whether the original gap is shrinking, and [Products](/features/products) to validate the exact catalog rows that were affected.
</Tip>

## Next steps

<CardGroup cols={2}>
  <Card title="Products" icon="package" href="/features/products">
    Inspect the specific catalog rows tied to each action.
  </Card>

  <Card title="Enrichment" icon="sparkles" href="/features/enrichment">
    Generate AI proposals for catalog, content, and metadata improvements.
  </Card>

  <Card title="Analysis" icon="chart-bar" href="/features/analysis">
    Understand the root cause behind your Actions queue before prioritizing.
  </Card>

  <Card title="Overview" icon="layout-dashboard" href="/features/overview">
    Monitor metric movement after you implement fixes.
  </Card>
</CardGroup>