Dashboard overview

The dashboard is your control center for AI visibility. After each audit, it automatically displays all the key metrics about how AI models perceive your brand.

It is divided into several sections:

6 key metrics — your visibility health indicators at a glance
Native vs Web comparison — how you perform on desktop AI vs web-augmented AI
Model breakdown — detailed performance for each AI model tested
Evolution chart — track your progress across multiple audits
GEO Checklist — website optimization score for AI visibility
Action plans — concrete steps to improve your scores

Tip: Each metric card has a ? button that explains what it measures. Hover or click on it for a quick explanation directly on the dashboard.

The 6 key metrics

Your dashboard displays 6 main metrics, each visualized as a donut gauge with a percentage or score. Together, they paint a complete picture of your AI visibility.

Visibility Score

Your overall presence in AI responses. Measures how often AI models mention your brand across all prompts. 100% = mentioned in every response.

Mention Rate

Frequency of mentions across all AI models tested. A high rate means multiple AI models consistently recommend you.

AI Share of Voice

Your market share in AI-generated responses compared to competitors. Shows how dominant your brand is in AI conversations about your industry.

Average Position

Your average ranking when AI lists recommendations. Position #1 = you are the top recommendation. Lower is better.

Sentiment Index

Measures the tone of AI mentions (0-10 scale). 10 = extremely positive. Captures whether AI models speak of your brand favorably or critically.

Topic Coverage

Breadth of topics where your brand appears. 100% = you are mentioned across all audit themes (products, services, comparisons, etc.).

Delta indicators: When you have multiple audits, each metric shows a +green arrow (improvement) or -red arrow (decline) compared to the previous audit. A small sparkline chart also shows the historical trend.

How scoring works

Each AI model's response is analyzed and scored on a 0-10 scale. The score depends on whether and how your brand is mentioned in the response.

Score scale

0 Absent

2 Position 5+

4-5 Position 4-5

6-7 Top 3

8 Position 2

10 #1 Pick

Score adjustments

The base score can be adjusted with bonuses and penalties:

+1 point — Detailed description of your brand (features, benefits)
+1 point — Recent or up-to-date information mentioned
+1 point — Positive tone or recommendation
-2 points — Negative sentiment or critical mention

Smoothing: AILabsAudit uses historical smoothing to reduce score volatility between audits. Scores are weighted against previous audit results to give more stable and reliable metrics over time.

Native vs Web scores

Your dashboard compares two types of AI responses side by side:

	Native AI	Web AI
What it is	AI responds from its training data only (no internet search)	AI searches the web before answering (real-time data)
Examples	ChatGPT (desktop), Claude (app), Gemini (app)	Perplexity, ChatGPT with browsing, SearchGPT
Depends on	Brand reputation in training data	GEO, online presence, recent content
Updates	Only when the model is retrained	Real-time with web content

Understanding the gap

The dashboard shows both scores with a VS indicator between them. Smart alerts help you interpret the gap:

Big gap (>20%) — Your brand relies too heavily on web results. Work on building your reputation in AI training data.
Medium gap (10-20%) — Web gives a noticeable boost. Keep investing in online presence.
Small gap, both high (>70%) — Excellent! Your brand is strong in both contexts.
Small gap, both low (<40%) — Priority: improve your overall visibility on all fronts.

Model comparison

The Model Breakdown section shows a detailed table where each AI model tested in your audit is listed with its individual scores:

Score (/10) — Overall performance score for that model
Sentiment — How positive or negative the model speaks about you
Position — Average ranking in that model's recommendations
Mention Rate — How often that model mentions your brand

Scores are color-coded for quick reading:

Green (7+) — High performance, the model knows and recommends you
Yellow (4-7) — Medium performance, room for improvement
Red (<4) — Low performance, the model barely mentions you

Radar chart

Next to the table, a radar chart provides a visual comparison of all models across multiple dimensions. This makes it easy to spot which models are strong and which need attention.

Why it matters: Different AI models have different training data. A model might recommend your competitor while another recommends you. Understanding which models favor your brand helps you target your optimization efforts.

Evolution and history

When you have two or more audits for a client, the dashboard unlocks additional features:

Progress banner

At the top of the dashboard, a banner summarizes your progress since the last audit:

Improved — Number of metrics that went up
Declined — Number of metrics that went down
Stable — Number of metrics that stayed the same

It also highlights your best gain and worst loss to quickly identify what changed the most.

Evolution chart

A line chart tracks all 6 metrics over time, with each audit date as a data point. This lets you visualize long-term trends and see the impact of your optimization actions.

Best practice: Run audits regularly (weekly or monthly) to build a meaningful evolution chart. This is the best way to measure the ROI of your AI visibility optimization efforts.

Reading audit results

The Results page (accessible from the audit detail) shows every prompt that was sent to each AI model, along with the full responses and individual scores.

Structure of the results

Results are organized prompt by prompt. For each prompt, you see:

The prompt text — the question that was sent to the AI models
A grid of response cards — one card per AI model

What each response card shows

Model name with a colored icon
Score badge — color-coded (green/yellow/red) with the score out of 10
Response preview — truncated text of the AI's response
Sentiment and position — mini indicators for each
Full response — click to expand and read the complete response in a modal

How to use results: Look for patterns. If your brand is absent from a specific type of prompt (e.g., "best X in Y"), you know exactly which topic to focus on. If one model never mentions you, investigate what content or data that model is missing.

GEO Checklist (Website analysis)

The GEO Checklist analyzes your website to evaluate how well it is optimized for Generative Engine Optimization — making your site easy for AI models to crawl, understand and cite.

Overall GEO Score

Your site receives an overall score (0-100%) with a letter grade:

A (70%+) — Well optimized for AI crawling
B-C (40-70%) — Some improvements needed
D-F (<40%) — Significant work required

The 4 categories

The score is broken down into 4 categories, each with its own progress bar:

AI Crawlability — Can AI bots access and read your site?
Structured Data — Does your site use schema.org markup?
Content Quality — Is your content clear, comprehensive and authoritative?
Authority & Trust — Does your site show expertise and trustworthiness?

Simple vs Detailed view

You can toggle between two viewing modes:

Simple view — High-level overview with expandable items. Click any checklist item to see details and recommended actions.
Detailed view — Full checklist with all items visible, ideal for a thorough review or PDF export.

Export: You can export the GEO Checklist as a PDF report or print it directly from the dashboard using the buttons at the top of the section.

Advanced GEO Modules

In addition to the GEO Checklist, the dashboard displays 5 advanced analysis modules that run automatically during each audit. These modules provide deeper technical insights about your website's AI readiness and competitive positioning.

Module 1 — SSR & AI Crawlability

Checks whether your website is technically accessible to AI crawlers. The widget displays a global score and 4 sub-scores:

SSR — Is your site server-side rendered or does it rely on client-side JavaScript?
Robots.txt — Does your robots.txt allow AI bots (GPTBot, ClaudeBot, etc.)?
WAF — Does your firewall block AI crawlers?
llms.txt — Does your site include an llms.txt file for AI instructions?

If specific AI crawlers are blocked, they are listed in red below the widget.

Module 2 — Entity Health

Evaluates how well your brand is represented as a knowledge graph entity. Sub-scores include:

Wikidata — Does your entity exist on Wikidata?
Properties — How complete is your Wikidata entry (description, logo, URLs, etc.)?
Schema.org — Does your website include structured Organization/LocalBusiness markup?
sameAs — Are your social profiles and external identifiers linked via sameAs properties?

Module 3 — Mention vs Citation

Analyzes the AI responses from your audit to classify how your brand is referenced. Uses AI classification to categorize each response into:

Mentions — Your brand is named in passing, in a list
Citations — Your brand is cited as a source or reference
Recommendations — Your brand is explicitly recommended

The widget also shows a mention-to-citation ratio, which indicates how often simple mentions become actual citations or recommendations.

Module 4 — Citation Readiness

Evaluates whether your website content is structured in a way that makes it easy for AI models to cite. Displays 7 sub-scores:

Sources — External references and links cited on your pages
Statistics — Presence of numbers, data, and measurable claims
Experts — Author attribution and expert quotes
Factual density — Ratio of objective, fact-based sentences vs. first-person/promotional content
Paragraphs — Well-structured content with clear paragraphs
Questions (H2) — Headers phrased as questions (matching how users query AI)
Direct answers — Content that directly answers common questions

Tip: The "Factual density" sub-score is one of the most impactful. AI models prefer citing content that presents objective facts and data rather than first-person promotional language.

Module 5 — STS Detection

An experimental module that analyzes your competitors for signs of Search Trust Signals manipulation — techniques used to artificially boost AI visibility. It checks for hidden text, authority claims, and suspicious patterns.

The widget shows how many competitors were analyzed and flags any suspects. If no anomalies are detected, a green checkmark is displayed.

Data quality alerts

When issues are detected with competitor data during the audit, the dashboard displays data quality alerts. These alerts help you understand potential problems that may affect the accuracy of competitive analysis.

Types of alerts

Unreachable URL — A competitor's website could not be reached during the audit (HTTP error, timeout, etc.)
Missing URL — A competitor was configured without a website URL, limiting GEO analysis
Suspicious competitor — A competitor shows signs of potential STS manipulation

Each alert can be dismissed individually by clicking the X button. Alerts are color-coded by severity: red for errors, orange for warnings.

Action plans

Based on your audit results, the dashboard displays action plans — concrete steps you can take to improve your AI visibility scores.

How actions are organized

Each action item includes:

Priority level — High, Medium, or Low (color-coded dot)
Title and description — What needs to be done and why
Status — Pending, In Progress, or Completed
Responsible party — Who should handle this action
Time estimate — Approximate effort required

GEO-generated actions

Some actions are automatically generated from the GEO module analysis. For example, if Module 1 detects that your robots.txt blocks AI crawlers, an action will be created to fix it. If Module 2 finds that your Wikidata entity is missing, an action to create it will be added. These GEO-based actions complement the AI-generated recommendations.

Tip: Focus on high-priority actions first. These are the changes most likely to have a significant impact on your scores in the next audit.

Tips for improvement

Here are proven strategies to boost your AI visibility scores:

Strengthen your online presence

AI models learn from web content. The more authoritative content exists about your brand online (blog articles, press mentions, directory listings, reviews), the more likely AI models are to mention you.

Optimize for structured data

Add schema.org markup to your website (Organization, Product, FAQ, HowTo). Structured data helps AI models understand your brand, products and services in a machine-readable way.

Create comprehensive content

Write detailed, factual content about your expertise. AI models favor sources that provide thorough, well-structured information over superficial content.

Build authority signals

Get mentioned in industry publications, partner websites and authoritative directories. The more trusted sources reference your brand, the higher your AI visibility.

Monitor and iterate

Run audits regularly and track your evolution chart. Identify which actions had the most impact and double down on what works.

Remember: AI visibility is a long-term game. Improvements in your online presence can take weeks or months to be reflected in AI model training data. Be patient and consistent.

Understand your AI Visibility Dashboard

Table of Contents