Skip to main content

This FAQ covers the questions we see most often when teams are getting started and when they are measuring results week over week. If you are troubleshooting missing data or unexpected results, also check Troubleshooting.

Getting started

What should we track first?

Start with 30–100 prompts that represent real discovery and evaluation:
  • “best __ for __”
  • “__ vs __”
  • “top ___ platforms”
  • “___ for beginners”
Then add 3–8 true competitors so comparisons are meaningful.

Why do my results look different day to day?

AI answers naturally fluctuate due to changes in sources, ranking behavior, and prompt interpretation. Use week-over-week trends and “Prev. Period” comparisons rather than single runs to make decisions.

What is the fastest way to get value?

A reliable early loop is:
  1. investigate one important prompt in Analytics → Prompts (open a losing run),
  2. generate one high-impact opportunity,
  3. fix one key page in Website → Pages.

Metrics and interpretation

What does Visibility mean?

Visibility is how often you are mentioned across runs for your tracked prompts and filters. Low visibility means you are frequently not included in answers for the queries you care about.

What does Prominence mean?

Prominence is your average rank when you are mentioned. Lower is better. If visibility is stable but prominence worsens, you are still included, but competitors are being preferred.

What does Share of Voice mean?

Share of voice estimates how much of the “recommendation space” you occupy relative to competitors across runs. It is best used directionally because answer formats vary.

What does Citation Rate mean in Meridian?

Citation Rate measures how often your owned domain is cited in answers (not just whether an answer has any citations). If citation rate is low, common next steps are:
  • improve citeable source pages (Website → Pages),
  • earn inclusion on high-impact sources (Outreach/Engage),
  • publish content aligned to what AI already trusts.

What does Sentiment measure?

Sentiment summarizes how AI describes your brand across dimensions like fees, reliability, and security. If sentiment drops, drill into the dimension and read the underlying negative mentions to understand the narrative.

Analytics questions

Why am I “Not mentioned” on a prompt?

Common reasons:
  • You do not have content that answers the query clearly.
  • Competitors are cited more often for that query.
  • Your site pages are not structured to be cited (FAQ/schema, clear headings).
  • The prompt set does not reflect the category queries where you are relevant.
The best next step is to open a losing run and inspect the citations and criteria used in the response.

Why is one platform worse than another?

Platforms rely on different sources and retrieval behaviors. It is normal to be stronger on one platform than another. Use Analytics → Citations and run details to learn which sources dominate on the weaker platform.

Why do products show “Not mentioned”?

Products are monitored against your tracked prompt set. If you are not tracking prompts where products could be recommended (for example “best basketball shoes”), products will often show Not mentioned. A good setup is to create a topic per product line/category and track at least 10 prompts per topic.

Website questions

Why don’t I see Referrals data?

Referrals requires a connected Google Analytics property. If you have connected GA but haven’t selected a property, you’ll see an empty state prompting selection.

Is Referrals the full amount of AI-driven traffic?

No. Referrals measures click-through sessions from AI platform referrers in GA. It will miss “view-through” influence where a user is influenced by AI but returns via another channel later. Use it as a directional trend indicator, not a complete ROI number.

Why is Crawlers empty?

Crawlers currently requires a Vercel integration. If you have not connected Vercel, you will see an empty state prompting connection.

Do crawler visits guarantee citations?

No. Crawling is often a prerequisite for discovery, but citations also depend on page quality, structure, and whether the ecosystem trusts your pages. Use Website → Pages to improve citeability.

Action questions (Opportunities, Content, Off-page)

What is the difference between Opportunities and Content?

Opportunities is the “what should we do next?” list. Content is where briefs and articles live. Many briefs start from Opportunities, but you can also create a brief directly in Content.

Why does Off-page matter if our website content is strong?

Many “best X” prompts are sourced from third-party editorial and community pages. If you are missing from those sources, you can be cited and ranked less often even with strong owned content. For Off-page, the most important thing is that credible, accurate language about your brand appears on sources AI trusts. A link can help, but it is often less important than the text itself being present in a trusted source.

Does adding our name guarantee visibility increases?

No. Many factors matter: whether the page stays in the citation ecosystem, whether your mention aligns with response criteria, whether competitors still dominate sources, and whether your owned pages are citeable.

Settings questions

What are “search terms” and why do they matter?

Search terms help Meridian detect mentions of your brand and common variations. If search terms are too narrow, you can miss mentions; if too broad, you can over-count. Review them in Settings → Company Profiles.

How do company profiles relate to data?

Company Profiles are separate datasets. Integrations and search terms are profile-specific. Make sure you are viewing the correct profile when analyzing results.

Who can change roles and settings?

Roles are typically restricted. Superadmins usually manage org-wide settings; Admins manage day-to-day; Members have limited permissions (exact permissions can vary).

Still stuck?

If you can’t find the issue here: