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

# Recommendations & actions

> The ranked plan that turns your GEO findings into work you can assign.

Two tabs turn analysis into a plan. **Recommendations** lists what to change. **Actions** sequences it.

## Recommendations

Every recommendation carries:

* **Impact** — Critical, High, Medium or Low.
* **Expected improvement** — how much closing it should move your GEO score.
* The **prompts** and **platforms** it affects.
* The **competitor** currently winning, where relevant.
* Concrete guidance on what to change.

The highest-impact recommendations also appear on the overview as **Priority fixes**.

### Categories

| Category                 | Typical work                                                    |
| ------------------------ | --------------------------------------------------------------- |
| **Content Gaps**         | Publish or expand a page that answers a prompt directly         |
| **Entity / schema work** | Structured data and entity markup so models can identify you    |
| **Publish the fact**     | State a claim explicitly that models currently have to infer    |
| **Pitch story**          | Earn a third-party mention on a source the models already trust |
| **Claim listing**        | Create or claim a knowledge-graph or directory entry            |

<Tip>
  *Publish the fact* recommendations are consistently the cheapest wins. They usually mean adding a sentence that states plainly something your site currently implies — a price, a location, a supported integration.
</Tip>

## Actions

The Actions tab is the sequenced plan: top priorities first, then the longer-horizon work.

| Grouping               | Meaning                                                            |
| ---------------------- | ------------------------------------------------------------------ |
| **High priority**      | Do these now; largest expected score movement                      |
| **Long-term strategy** | Structural work — authority, entity presence, third-party coverage |
| **No action**          | Findings that don't warrant work right now                         |

Actions are marked **Fixed** once addressed, so the plan reflects real progress rather than resetting each run.

## A working rhythm

<Steps>
  <Step title="Run the analysis">
    Weekly or monthly, depending on how fast you publish.
  </Step>

  <Step title="Take the top priorities">
    Pull the high-priority actions into whatever tracker your team already uses.
  </Step>

  <Step title="Ship them">
    Content gaps go to the [Blog Agent](/content/blog-agent); on-page issues go to the [Page Agent](/content/page-agent); technical items go to [Site Audit](/search/site-audit).
  </Step>

  <Step title="Re-run and compare">
    Compare against the previous report to see exactly what your work moved.
  </Step>
</Steps>

<Note>
  AI answers shift on their own as models update. Some movement between runs is the models changing, not your work — which is why comparing consecutive runs matters more than reading any single one.
</Note>

## Next steps

<CardGroup cols={2}>
  <Card title="Blog Agent" icon="pen-nib" href="/content/blog-agent">
    Generate the content these recommendations call for.
  </Card>

  <Card title="Reports" icon="file-lines" href="/workspace/reports">
    Compare runs and prove the impact.
  </Card>
</CardGroup>