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

# Your first 30 days of GEO

> A week-by-week plan from first scan to the first prompts you win.

This is the order we recommend for a new project. Each week builds on data from the one before, so don't skip ahead to writing content.

## Week 1: Make the data trustworthy

<Steps>
  <Step title="Check how engines name you">
    Open a few answers on the [Prompts](/docs/geo/prompts) page and look for product names, abbreviations and your bare domain. Add every variant as an [alias](/docs/geo/brand-and-domains), because missing aliases are the most common reason a mention rate looks too low.
  </Step>

  <Step title="Complete the competitor list">
    Share of voice and content gaps only see [tracked competitors](/docs/geo/competitors), so import the full list by CSV and add synonyms.
  </Step>

  <Step title="Tune prompts">
    Pause prompts that name your brand or are too broad to ever mention you. Add the questions your sales team hears, and see [Writing good prompts](/docs/geo/prompts#writing-good-prompts).
  </Step>

  <Step title="Pick engines and languages">
    Match [engines](/docs/geo/engines) and [languages](/docs/geo/languages) to where your buyers are, since every engine and language multiplies the answers a scan uses.
  </Step>

  <Step title="Install the tracker">
    Add `@usenotra/geo` to your site so you can see AI crawlers and referrals from day one. See [AI traffic](/docs/ai-traffic/overview) for setup.
  </Step>
</Steps>

## Week 2: Find where you lose

* Read [Content gaps](/docs/geo/content-gaps), sorted by opportunity.
* Open [Shelf space](/docs/geo/shelf-space) and note the third-party pages where competitors are listed and you are not.
* Run [Agent readiness](/docs/geo/agent-readiness) and fix the **Must do** items.

Then pick five targets, mixing prompts to answer with your own content and pages to get listed on.

## Week 3: Ship

* Plan briefs for your top gaps with the [GEO writer](/docs/geo/writer) and publish the drafts. Before publishing, add statistics, quotes and cited sources, which in the [GEO study](https://arxiv.org/abs/2311.09735) raised visibility in AI answers by 28 to 43%.
* Update existing pages instead of only writing new ones. [Seer Interactive](https://www.seerinteractive.com/insights/study-content-recencys-impact-on-ai-visibility-in-2026) found 75% of AI-cited pages were updated within the last year.
* Open shelf tickets, assign them and set due dates.
* Make sure new pages are served to AI agents by checking the [AI traffic](/docs/ai-traffic/overview) page for crawler visits on them.

## Week 4: Measure

* Use [What changed](/docs/geo/troubleshooting/numbers-changed#what-changed) on the Overview page to see which prompts moved.
* Compare share of voice with week 1.
* Repeat with the next five targets.

<Tip>
  Engines that search the web (ChatGPT search, Perplexity, Google AI Overviews) can cite a new page once it is indexed, which takes [hours to weeks](https://developers.google.com/search/docs/appearance/ai-features). What a model knows from training data lags by months, so judge new content by crawler visits first and mentions second.
</Tip>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.