Quickstart guide

This quickstart walks you through the fastest path to value in Meridian:

1. Track the prompts that matter
2. Understand your visibility baseline
3. Generate your first brief to start improving visibility.

​

1) Add your first prompts

Prompts are the questions you want to win in AI assistants. The quality of your prompts determines how useful your dashboards will be. Go to Analytics → Prompts, review the initial prompts generated, and add 10–25 prompts to start.

​

What good prompts look like

Use conversational questions that match how people ask AI tools for recommendations:

• “best __ for __”
• “top ___ tools/platforms”
• “__ vs __”
• “___ for beginners”
• “how do I choose ___”
• “is ___ worth it”

​

What to avoid (for now)

• Purely branded prompts only (e.g., “What is trymeridian.com?”). These are helpful later, but they won’t tell you whether you’re being discovered.
• Prompts extremely specific to your brand (e.g., “What is the best GEO platform based in New York that starts with the letter ‘M’)
  Adding prompts that you already have high visibility means we’ll only be trying to increase visibility in queries that you already appear for.
• Prompts that are too broad (e.g., “crypto” or “marketing”). Broad prompts are hard to act on because the answer criteria varies a lot.

​

2) Set your competitor context

Meridian is comparative by nature because AI answers often return lists. Adding competitors helps you understand who is being recommended alongside you (or instead of you). Go to Analytics → Competitors and add 3–8 true substitutes.

​

Tips for competitor selection

• Pick companies your customers would genuinely consider instead of you.
• You can always adjust this later as you learn who shows up in answers.

​

3) Check Home to establish your baseline

Home gives you a quick read on:

• Visibility: how often you’re mentioned across your tracked prompts
• Citation Rate: how often answers include sources
• Sentiment: how AI describes you across decision-driving dimensions
• Prominence: your average rank when you’re mentioned

​

What to do if your baseline looks low

That’s the best starting point. We want to add prompts where you want to increase visibility in, instead of solely prompts you’re already showing up for. Low baseline usually means one of:

• You’re not included in answers yet (coverage/authority gap),
• Competitors have stronger source presence (citation gap),
• Or your website/content isn’t structured to be cited (site readiness gap).

You’ll address these in the next steps.

​

4) Investigate one important prompt

Go to Analytics → Prompts and pick a prompt that matters to your business. Then:

1. Open the prompt’s analysis (press Analyze)
2. Find a run where you were not mentioned or ranked low
3. Open the run details to see:
   • Who outranked you,
   • What criteria the answer used,
   • And which sources were cited

This gives you immediate clarity on what is driving your current rank: sources, structure, narrative, or product coverage.

​

5) Add your Brand Kit (so outputs are accurate and on-brand)

Go to Brand Kit and fill out the basics before you generate content. Brand Kit is the “ground truth” Meridian uses when it creates briefs and articles. If it’s empty or vague, generated content tends to be generic. If it’s specific, your content is more likely to reflect the language and criteria AI uses to rank answers (and less likely to include risky or inaccurate claims). At minimum, add:

1. Positioning (1–2 sentences) Who you are, who you’re for, and what you do.
2. Key differentiators (with proof where possible) List 5–10 differentiators that matter for “best X” prompts (price, trust, support, speed, etc.). Include links to source pages if you have them.
3. Approved vs disallowed claims Write down what content can safely say and what it must not say. This is especially important in regulated or high-stakes categories.
4. Voice & style guidelines A few rules like “short sentences,” “use bullets,” “include decision criteria tables,” “avoid vague superlatives.”
5. (Optional) Writing samples Add 1–3 samples so Meridian can match your tone.

​

6) Generate your first brief (and optionally an article)

Go to Opportunities and pick one “High” priority recommendation. Click Generate brief. Then go to Content and open the brief. From there you can:

• Review the structure and target intent
• Generate an article draft
• Refine it
• And publish it in your website

​

7) Improve your website readiness (fast win)

Go to Website → Pages. This is where Meridian scores your site for AI discovery and citation-readiness. Start by scanning the top summary scores (Robots, Performance, Technical SEO, Content). These scores help you quickly see what’s limiting visibility; whether AI can crawl you, understand you, and confidently cite you. Next, scroll the page list and a page with an issues flag. Prioritize:

• A page that matters to your business (product/category/guide)
• A page with a higher issue count or lower Technical/Content score.

Click into that page to open its recommendations. You’ll see specific fixes (for example: add an FAQ section, add FAQ schema, fix missing meta description, or other technical/content improvements). From there you can:

• Review the recommended fix and why it matters
• Follow the step-by-step action steps Meridian provides
• Implement the change on your website (in your CMS or codebase)

If you only make one website change this week, start with a recommendation that improves citeability (like FAQ + schema) on a page that should be a “source page” for your category. This is one of the most reliable ways to increase owned citations over time and improve rankings.

​

8) Capture quick wins with Off-page (Editorial, Directories, Social)

Go to Off-page and pick one surface to work on this week. Off-page is where you earn mentions in the sources AI already trusts for your prompts. A good default is to spend 15–30 minutes and pick a small set of targets you can actually complete.

​

Editorial (lists and articles)

Go to Off-page → Editorial and pick 3–5 high-impact targets. Editorial is a list of third-party articles that are actively being cited for your tracked prompts. If competitors are included and you are missing, requesting inclusion can be a fast way to improve the source ecosystem AI pulls from. When you reach out, include:

• who you are and who you’re for,
• why you belong in the list,
• what differentiator matches the criteria,
• and a credible source page that supports the claim.

​

Directories (structured listing sources)

Go to Off-page → Directories and pick 3–5 directory targets. Directories can be high leverage because listings are structured and easy for AI to summarize. When you create or update a listing, focus on adding specific, on-brand language:

• what you do and who you’re for,
• your differentiators (with proof where possible),
• and “best for” positioning that matches how people choose in your category.

​

Social (UGC and community)

Go to Off-page → Social and pick 2–3 threads or community pages. In some industries (often DTC), AI relies heavily on community sources. Social is less about links and more about trusted language showing up in places AI already cites. Teams use Social in two ways:

1. participate in existing threads that AI is actively citing for your prompts, and
2. use those threads as inspiration for what to post next (and where) on your own social channels.

​

9. Import your products for SKU-level visibility (For Ecommerce only)

If you sell physical products that can appear in shopping-style AI results (product cards), you can import your catalog so Meridian can track product-level outcomes. Go to Analytics → Products and connect your store to import SKUs (currently Shopify). Once your catalog is connected:

• Products shows which SKUs are being mentioned vs not mentioned for your tracked prompts.
• You can click See responses to understand which prompts trigger recommendations, what criteria the answer uses, and which competitors appear instead.