AI-powered search is transforming the way people discover information. Platforms like ChatGPT, Google Gemini, and Perplexity now generate direct answers instead of showing only traditional search results.
Note: This is only about LLM tracking. AI Overviews (AIO/SGE) are detailed in the dedicated article.
How It Works
AI Search Tracking allows you to see:
AI Search Presence: the percentage of Search Volume where the site is mentioned, calculated from the total search volume of keywords in this group that trigger AI to mention brands.
AI Search Cited: the percentage of Search Volume where the site is cited.
Citation Rank: the position among AI-generated citations (#1, #2, etc.).
Context Analysis & Sentiment: whether the brand is mentioned positively, negatively, or neutrally.
AI Response Examples: actual snippets of how AI presents the site’s content.
Why this matters:
Mentions reflect how often the model recommends your brand within the narrative.
Citations reveal the evidence the model leans on (your pages and credible third-party sources).
At topic (keyword) level, this tells you where you have authority and where you need deeper content, stronger topical coverage, or better citations.
How SEOmonitor queries AI models
SEOmonitor currently supports three AI models:
ChatGPT (OpenAI)
Gemini (Google)
Perplexity
For each keyword in your campaign:
SEOmonitor sends the exact tracked keyword to the selected AI model(s) in a fresh, neutral session (no user history, no personalisation).
If the model asks for clarification, SEOmonitor provides short, neutral brand context so the answer is complete and comparable.
The system then detects:
Whether your brand is mentioned (and sentiment)
Whether your site is cited
Which other brands and sources are mentioned or cited
Why this approach:
LLMs do not expose real prompt volumes (there is no “GSC for LLMs”), so any “prompt demand” would be synthetic.
Keywords already represent the underlying lookups LLMs run in the background when they go to search engines to collect sources.
Using the same keywords as for Google ranking keeps AI Search tracking standardised and comparable across brands and campaigns.
Multi-LLM Tracking
You can now track multiple AI models simultaneously on the same campaign. Pro plan users can purchase and activate ChatGPT, Gemini, and Perplexity independently, each building its own historical data from the date it is enabled.
Key points:
Simultaneous tracking: No need to choose just one — enable any combination of the three models.
Independent data: Each LLM stores and preserves its own historical data. Switching your view between LLMs does not reset or affect data for any model.
Processing priority: Keywords are processed across all active LLMs in this order: ChatGPT → Gemini → Perplexity. A keyword is marked “complete” only once it has been processed by all active models.
One-time Snapshot vs. Ongoing Tracking
AI Search Snapshots are free:
One-time processing of all campaign keywords.
Includes your AI presence, citations, and basic context.
Limitations: no history, no exports, no automatic updates.
One snapshot per campaign. The snapshot is retained until you purchase an LLM add-on.
AI Search Tracking comes as a paid add-on:
Weekly automatic updates.
Historical trend charts and export options.
Continuous monitoring of your and competitors presence.
Integration with your existing keyword tracking.
Getting Started
1. Run your first snapshot
Open the campaign in Rank Tracker and click "Get free snapshot"
Instantly see your presence, citation ranks, and context.
Review your brand terms: SEOmonitor will suggest detected terms, and you can confirm, add, or remove them before the snapshot runs.
Note: Any brand term changes made later (in Campaign Settings) will take effect after the next data refresh (up to 24 hours).
2. Enable ongoing tracking
Turn on weekly tracking for historical trends, exports, and continuous monitoring.
When you first activate AI Search or add new keywords, these go through a separate async processing. During this time, you’ll see progress indicators until the fresh data becomes available.
Note: Processing time varies depending on your keyword volume and the overall queue. Large campaigns and busy periods may take longer to complete.
Some keywords may be automatically excluded if they can’t return meaningful results (e.g., no AI citations or no search results). Excluded keywords are flagged accordingly in your list and do not count toward billing.
3. Add more LLMs
You can add additional AI models at any time:
From the LLM Switcher in Rank Tracker — click on a greyed-out (unpurchased) LLM icon to trigger the purchase modal.chase modal.
From Campaign Settings — navigate to the AI Search add-ons section and select the models you want to enable.
When you add a new LLM:
Data tracking begins immediately after enabling new LLM.
Your existing data for other LLMs is not affected.
A confirmation modal will show the pricing impact before you confirm.
The LLM Switcher
The LLM Switcher is a new component in the Rank Tracker interface, located next to the device selectors (Desktop/Mobile).
How it works:
Active LLMs display their icons (ChatGPT, Gemini, Perplexity) in full colour.
Unpurchased LLMs appear greyed out with a purchase trigger — click to open the purchase modal.
Default selection: ChatGPT is selected by default. In subsequent sessions, the switcher remembers your last-selected LLM.
Instant switching: Toggling between purchased LLMs updates all dashboard widgets, charts, tables, and the Competition tab instantly — no page reload required.
What updates when you switch:
Component | Behaviour |
Top stats & scorecards | Refresh to show metrics for the selected LLM |
Trend charts | Update trend lines, percentages, and metrics per selected LLM |
Keyword table | Update columns for citations, mentions, and sentiment per LLM |
Competition tab | Show competitor performance, trends, and Share of Voice for the selected LLM |
Processing status (top bar) | Show processing status for the currently selected LLM |
Processing status (footer) | Show total keywords pending across all active LLMs |
Data for Performance reporting:
This section explains the data model, frequency, and where to find the most useful charts and views for reporting.
Weekly update cadence
AI Search visibility updates once per week (per crawl).
Charts show one data point per crawl.
If you select dates between crawls, the line will simply jump from one crawl to the next; there is no interpolation.
Trends and percentage deltas are calculated between actual crawl points.
New keywords added between crawls appear on the previous crawl date once they have data.
Data gaps can occur at the beginning or end of a selected time range if the dates do not align with specific crawl days. To ensure accuracy, align both start and end dates with crawl schedules when analyzing rankings over time.
Where to see AI Search reporting data
You can surface AI Search performance for your domain in two main places:
Scorecards
In the Rank Tracker Strategy view, check AI Search Mentioned card for:
AI Search Mentioned
AI Presence Sentiment
AI Search Cited
AI Search Citation Rank
These represent your domain's visibility for AI Search for the currently selected LLM. For competitors AI Search mentions and citation share switch to the Competition view.
Charts and Trends
Review how your brand’s performance in AI Search is changing WoW through dedicated charts and compare this with organic and AIO metrics. All chart data reflects the LLM currently selected in the switcher.
Competitive Analysis & Reporting
Use the Competitors panel to compare how your performance compares against competitors for Google, AI Overview, and AI Search.
What each column means
For each competitor and for the same keyword set:
Google
Organic visibility and rank metrics (as usual).
AI Overview (AIO)
Mentioned: % of search volume where the brand is named.
Cited: % of search volume where the site is linked in AIO.
AI Search
Mentioned: % of search volume where the brand is named in the LLM answer.
Cited: % of search volume where the site is cited as a source.
Deltas show the change vs the previous crawl. All AI Search metrics in the Competitors panel are specific to the LLM selected in the switcher.
Understanding Discrepancies in Metrics
Differences in visibility metrics and competitor percentages across campaigns can arise from non-deterministic AI responses. Since AI outputs are not always consistent, responses for the same keywords and dates may vary between campaigns. Recognize this as a normal characteristic of AI systems.
All of these rates are weighted by search volume, so wins on high-volume topics matter more than lower Search Volume topics.
When comparing results across LLMs: each model has its own knowledge base, citation preferences, and response patterns. Differences between models are expected and informative — they show where each model sees your brand’s authority differently.
AI Share of Voice measures your share of influence amongst competitors selected on an 'Impression Score' for each keyword.
How the Score is Calculated:
- If your brand is mentioned in AI Search, you get 1pt for a mention or
0.5pts for a citation.
- If no brands are mentioned, you get 1pt for a citation.
The same calculation is used to calculate the score for every competitor you select.
Scores are multiplied by search volume, and your Share of Voice is your
share of the total score among all analyzed competitors.
How to apply AI Search insights in reports and analysis
1. Start at the overall level
In All keywords, review:
AI Search Mentioned %
AI Search Cited %
Average Citation Rank
Sentiment distribution (positive / negative / neutral)
Compare with:
Google visibility
AIO metrics
Questions to answer:
Is AI Search visibility broadly aligned with organic visibility, or lagging?
Is sentiment mostly positive or neutral?
Are your citations in line with your mentions, or are you “mentioned without being cited”?
2. Benchmark against competitors
Go to the Competitors panel (All keywords).
Compare your AI Search Mentioned/Cited against key competitors.
Use this to label your situation:
You lead in both mentions and citations.
You lead in citations but lag in mentions.
You lag in both.
This helps you position AI Search in reporting:
“We are strong in citations but behind in pure brand visibility.”
“Competitor X is gaining SoV in AI Search for our core topics.”
“We gained AI Search visibility while Google visibility stayed flat.”
3.Group drill-down
Use your existing keyword groups (e.g. product lines, themes, funnel stages) to see where LLMs see you as an authority.
Select groups in the sidebar
Compare for that group:
AI Search Mentioned / Cited
Google visibility
AIO Mentioned / Cited
Look for:
Groups where AI Search visibility is higher than organic:
The model already recognises you as a strong source. Protect and expand.
Groups where AI Search visibility is lower than organic:
You rank well in Google but are underrepresented in LLM answers.
These are prime candidates for optimisation (content depth, citations, PR).
You can also open the group-level Competitors view to see who dominates AI Search for that topic cluster.
4. Keyword-level drill-down & gap analysis
From the keyword table for any group or for all keywords:
Use filters to find strategic sets:
Examples:
“We are absent, competitors present”
Filter:
Only other brands mentioned
Sort keywords by:
Search volume (highest first)
Difficulty (start with “easy” or “medium” topics)
These are high-impact gaps where LLMs already talk about competitors, and you have topical authority.
“Negative sentiment”
Filter where:
Your brand is mentioned negatively
Review the AI response and cited sources to understand why.
5. Read AI responses and citations for content ideas
Once you have a filtered keyword list:
Open the AI response for the top topics.
Check:
How your competitors are described.
What features, benefits, or content formats the LLM highlights.
Whether the response is purely informational or brand-led.
Switch to the Citations panel:
Identify which sites are cited besides you (retailers, publishers, comparisons, guides, blogs, etc.).
Flag high-authority third-party sites repeatedly cited across important topics.
Use this for:
On-site content planning
Add missing content types (guides, comparisons, FAQs, styling tips, etc.).
Make pages more complete and aligned with what the LLM emphasises.
Digital PR & partnerships
Target the repeatedly cited third-party domains as priority outreach targets.
Aim to earn mentions and links where competitors are already present.
Tip! Use the API for bulk content analysis
6. Use trends to validate impact
After you implement changes:
Track AI Search Mentioned and AI Search Cited over the next weeks.
Check if the affected groups or keywords show:
Increased mentions for your brand.
More citations in the AI response.
Monitor competitor curves on the same trend chart to see if gains come from:
Market growth, or
Share taken from specific competitors
Practical Tips for Interpreting Discrepancies
Align data analysis with crawl schedules to reduce gaps in analysis.
Understand that AI outputs often vary between campaigns due to inherent non-deterministic behavior.
Focus on long-term trends rather than isolated variations to glean meaningful insights.
Compare across LLMs — Use the switcher to review the same keyword set across ChatGPT, Gemini, and Perplexity. Differences highlight where each model’s training data or retrieval preferences create distinct opportunities.
Billing & availability
Pricing — Charged per 1,000 keywords (rounded up), per active LLM per month. Each 1,000-keyword bucket costs €19 per LLM provider.
Keyword stacking — If you track 1,000 keywords across 3 LLMs, that counts as 3 × 1,000 = 3,000 billed keyword units.
Trials — AI Search is enabled for the first campaign (up to 10,000 keywords). The free snapshot uses ChatGPT by default.
Compatibility — Available for Pro plans and above.
Not available — Legacy / grandfathered plans and Content Writer–only plans require an upgrade. Grandfathered plan users can see LLM icons and view snapshots, but are prompted to upgrade via a banner.
Notes:
Free snapshots reserve tracking slots for the entire billing cycle; archiving campaigns or keywords does not free them mid-cycle.
You can choose which campaigns to enable the add-on for and roll out gradually across your client base.
Adding or removing LLMs is handled via Campaign Settings or the purchase modal, with confirmation steps showing pricing impact.
Removing an LLM stops weekly updates for that model and reduces your billing accordingly.
Limitations & known constraints
Draft campaigns — AI Search is not supported on draft (free) campaigns, since processing costs are higher than standard Google tracking.
Snapshots — Require regular Google tracking to be active and do not include exports or history. One snapshot per campaign; retained until you purchase an LLM add-on.
Trial plans — Only ChatGPT is available. Other LLMs are not accessible during the trial period.
Starter plans — Only the free AI Search Snapshot is available (no ongoing tracking add-on). Upgrade to a Pro plan to enable weekly tracking and multi-LLM support.
Other modules — AI Search will be integrated later into Research and Forecast, following Organic Traffic enhancements.
Non-deterministic results — AI models may produce different responses for the same keyword across different sessions. This is expected behaviour and does not indicate a system error.
Last updated: March 11, 2026