CitadexDocs

Dashboard: how to read it and use it

The Dashboard is the first page you open each day: the AI visibility score, mentions / citations / share of voice, this week’s tasks, the path from mention to recommendation, and breakdowns by country and engine. This manual walks through each area and what to click.

Product Manuals8 min read

What this page is for

The Dashboard summarizes every scan on one page so you can see whether your content and changes are increasing AI recommendations: a visibility score, mention and citation counts, the tasks worth doing first this week, how often a mention turns into a recommendation, and which country and which engine are weakest.

When to use it

  • First thing each day, to check quickly whether anything has changed.
  • For a weekly meeting or a client update: use Export PDF.
  • When the data looks unusual: click a number and check the evidence.

The page, top to bottom

Header: the title reads AI Visibility: yourdomain. On the right: the Prompts pill (how many prompts you track), How to use (a 6-step workflow, each step with Go), the Export PDF menu, and the gear icon Report branding.

Filter bar: Global shows all markets together; each country chip filters to one market. Add market opens the add-market dialog. Platforms selects AI engines. On the right is the date range (Today / Last 7 days / Last 30 days / Last 3 months / Last 6 months / All time). Engines your plan does not include are greyed out with a lock.

  1. Key metrics + AI Visibility gauge: the left card has the tabs Key metrics, Mentions and Citations; the three numbers are Mentions, Citations and Share of voice. Click a number to open the evidence drawer and see which prompts and engines produced it. The gauge on the right is a 0–100 visibility score labelled Low · needs work, Medium · climbing or High · strong, with Platforms, Avg. position and Positive % underneath.
  2. This week’s tasks: up to 5 tasks, comparing the last 14 days with the 14 before. Each task has a type tag: Recommended less, Citations lost, Competitor ahead, Prompt gap, Page not cited, Cited, recommends others or Wrong brand info. Cited, recommends others means AI cites one of your pages but its answer recommends other brands from it (common with “N best” comparison posts); open the change list to name your brand in the first sentence and put it first in the comparison. Tasks are ranked by impact and business value, and also by topic revenue once Revenue Attribution is connected. When an engine changed for everyone in the same period, the task says so.
  3. From mention to recommendation: AI answers → Mention you → Cite your site → Recommend you → First pick, overall and per AI engine. It also shows the share of answers citing you that recommend you, the Weighted recommendation share (a recommendation alongside 3 tracked competitors counts 1/4), and Sole recommendations. Rates show a 95% range; the Data source line says whether answers were captured from the real product UI or the official API, and which models answered. Being cited is not the same as being recommended; this card shows the gap.
  4. Platform-wide changes: appears only when an AI engine changed for all projects at once, for example fewer cited links per answer, or a new model in use. The same days are marked on prompt trend charts.
  5. Mentions by country/region and Coverage by AI platform: the left card lists mention rate per market, with Add country market at the bottom; the right card switches between Mentions and Cited pages per engine.
  6. Coverage by query language and Language visibility gap: shown only when there is enough data and one language is clearly weaker. It lists the prompts you are absent from and the competitors that appear there.
  7. Performing Queries: filter by Brand / Competitor / City / General keyword, search, sort by visibility or mentions, and filter by intent and visibility band. Monitor at the end of a row opens Prompt Tracking.
  8. Query log: every single check, filterable by model, appeared / not appeared and type. View opens the evidence.
  9. Recommended actions: three cards — Find trending topics, Find competitor gaps and Check your domain's AI health.
Dashboard
AI Visibility
42 /100
Mentions
38
Citations
17
Share of voice
23%
👆 Click Mentions / Citations / Share of voice to see the evidence

Common tasks

Work through this week’s tasks

  1. Read the task line: the prompt or page, what changed, and the buying stage.
  2. See evidence shows the answers behind it.
  3. Act: Write article (for a prompt gap or a competitor ahead) opens the writer with the prompt filled in; credits are confirmed there before anything runs. Open change list opens the page’s change list for a page task.
  4. Click ✓ when you have done it. The task stays as Done and moves to Resolved once the problem is gone. ✕ dismisses a task; Undo brings it back. Viewers see the tasks without ✓ and ✕.

See where a number comes from

  1. Click Mentions, Citations or Share of voice.
  2. The evidence drawer lists the prompts and engines behind it. The share-of-voice view shows your mentions, competitor mentions and the most-mentioned competitors.
  3. For the full table, go to Prompt Tracking.

Export for a client or colleague

  1. Click Export PDF, then Preview report to check it, or Print / Save as PDF directly.
  2. Export data (CSV) downloads the raw table.
  3. On an account with white label, if branding is not set yet you will see Report branding not set: choose Use Citadex or Set up branding.

Add a market

Click Add market in the filter bar or Add country market at the bottom of the country card. Every market has a country, and engines answer from that country. In the Add market dialog, choose the country, language, an optional city and the scan engines, then click Add market. The dialog adds the market only, with no prompts, and adding a market uses no credits. If the country and language are already a market of the project, the dialog says “<market> is already a market of this project, so there is nothing to add” and points to Markets & engines for changing its engines. Find the new market's prompts in the Prompt Library with Find candidate questions, or paste them with Add Prompt; checks start once prompts are tracked.

Switch project

Use the project switcher at the top left. Dashboard, Prompt Tracking, Integrations and every other page show data for the selected project.

Peer benchmark

  • Peer benchmark on the Dashboard compares your mention, recommendation, citation and ad rates with other Citadex customers in the same sector and market: the peer median, the top-quarter mark and the share of peers you are ahead of.
  • Aggregates only: it appears once at least 5 different customers stand behind it and never shows another customer's brand, prompts or answers. While your sector has fewer than 5, it compares with every sector in that market.
  • Biggest gap names the AI engine to work on first; sites in Sites AI cites most in peers' answers marked not in your answers yet are worth getting mentioned on.
  • The sector is detected automatically; change it on the card if it is wrong.

Common questions

Why does it say "No AI visibility data yet"? The project has not been scanned. Click Run your first scan to start the new-project flow, or, for an existing project, add prompts in Prompt Tracking; new prompts are queued for checking automatically.

The numbers don’t match Prompt Tracking. The Dashboard follows the market / platform / date filters at the top, while Prompt Tracking shows each prompt’s latest result. Set the filters to Global and All time before comparing.

An engine is grey with a lock. Your plan does not include it. Free and Pro include the standard 7 engines (ChatGPT, Claude, Gemini, Perplexity, Copilot, Google AI Overviews, Google AI Mode); Business, Enterprise and agency plans include every engine, adding Grok and Mistral. China markets use Chinese engines such as Doubao, DeepSeek and Qwen, which Free and Business and above include and Pro does not.

Visibility dropped sharply. First check Platform-wide changes: if an engine changed for everyone on the same days, the drop is likely the engine, not your brand. Then check Coverage by AI platform for a single engine whose coverage fell, and the country card for a newly added market lowering the average.

Troubleshooting

  • The page stays on Loading your AI visibility data…: refresh once; if it persists, switch to another project and back.
  • Metrics could not be loaded: the server did not answer in time. Your data is intact; click Retry.
  • Every metric is 0 but Prompt Tracking has data: check whether the date range and market filter exclude it.
  • The exported PDF is blank: open Preview report, let it finish loading, then print.

Was this article helpful?