You Can't Improve What You Don't Measure
With Google, tracking visibility is straightforward. Check your rankings. Use Google Search Console. Monitor your traffic. But AI search is different—there are no rankings, no clickthrough rates, and no analytics dashboard.
Yet tracking is just as critical. If you're investing time and resources into building AI visibility, you need to know if it's working.
This guide shows you exactly how to track your AI presence, what metrics matter, and which tools can help. If you're just getting started, first read about [how AI search differs from Google](/learn/how-ai-search-differs-from-google) and follow our [step-by-step guide to getting into ChatGPT's knowledge base](/learn/step-by-step-chatgpt-visibility).
The Manual Testing Method
Before we get into tools, let's cover the basics: manual testing. This is free, immediate, and gives you direct insight into how AI tools present your business.
Setting Up Your Testing Protocol
Step 1: Create Your Query List
Write down 15-20 questions that should ideally mention your business. Include:
Direct queries:
Industry queries:
Problem-solving queries:
Step 2: Test Systematically
Test each query weekly in:
Step 3: Document Results
For each test, record:
Use a simple spreadsheet:
| Date | Tool | Query | Mentioned? | Position | Context | Others Mentioned |
|---|
What to Look For
Positive signals:
Negative signals:
Neutral signals (neither good nor bad):
Testing Frequency
Phase 1 (Months 1-3): Test monthly
Phase 2 (Months 4-6): Test bi-weekly
Phase 3 (Months 7+): Test weekly
What Manual Testing Reveals
Manual testing isn't just about whether you're mentioned—it reveals how AI understands your business.
Example findings you might discover:
"When I ask about 'best Italian restaurants in Denver', I'm never mentioned. But when I ask 'authentic Italian restaurants with outdoor seating in Denver', I show up second. This tells me AI associates me with authenticity and outdoor dining."
"ChatGPT knows my business name but gets my location wrong, saying I'm in Portland when I'm actually in Seattle. I need to fix location consistency."
"Claude mentions my business for enterprise clients but not small businesses, even though I serve both. My marketing must emphasize our work with smaller companies more."
These insights are actionable. They tell you what's working and what needs adjustment.
Tracking Third-Party Mentions
While you can't directly track what goes into AI training data, you can track the sources AI draws from.
Google Alerts
Set up Google Alerts for:
Configure alerts to:
Every alert is a potential source AI might reference.
Media Monitoring Tools
For more comprehensive tracking, consider:
Free Options:
Paid Options ($50-$200/month):
What to track:
Why This Matters for AI
Each mention is a data point AI might use. Tracking mentions helps you:
Tracking Authority Signals
Remember: AI visibility is about authority, not traffic. Track the signals that build authority. Learn more about [the specific authority signals that make LLMs trust your brand](/learn/authority-signals-llms-trust).
Verification and Certifications
Track when you gain:
Each of these increases AI's confidence in your legitimacy.
Press Coverage
Maintain a "press page" on your website listing:
Update this monthly. It serves double duty: shows visitors your credibility and gives you a quick reference for your authority buildup.
Speaking and Expert Contributions
Track:
Each appearance is evidence of expertise.
Reviews and Ratings
Monitor across platforms:
Use a spreadsheet:
| Platform | Jan | Feb | Mar | Apr | May | Jun |
|---|---|---|---|---|---|---|
| 23 | 27 | 31 | 34 | 39 | 45 | |
| Yelp | 12 | 13 | 14 | 16 | 18 | 21 |
Steady growth signals increasing legitimacy.
Using Search Console and Analytics
While Google Search Console won't tell you about AI, it reveals something valuable: how people find you now, which informs your AI strategy.
Check for:
If people are already finding you for certain topics on Google, those are prime targets for AI visibility too.
Analytics insights:
The Coming Soon: Automated LLM Monitoring
Manual testing works but it's time-consuming. The future is automated monitoring tools that:
Several companies (including us) are building these tools. Manual testing is your bridge until automated solutions are widely available.
What to look for in an LLM monitoring tool:
Creating a Monitoring Dashboard
Pull everything together in a simple dashboard. Use Google Sheets or Excel.
Tab 1: Manual Testing Results
Tab 2: Third-Party Mentions
Tab 3: Authority Signals
Tab 4: Reviews
Tab 5: Monthly Summary
Update monthly. Review quarterly. Adjust strategy based on what you see.
Red Flags to Watch For
Warning sign: Declining mentions
If you were being mentioned but now aren't, investigate:
Warning sign: Incorrect information
If AI mentions you but gets facts wrong:
Warning sign: Negative context
If you're mentioned but in negative ways:
Benchmarking Against Competitors
Don't just track yourself—track 3-5 key competitors.
Use the same queries but note:
Competitor analysis reveals:
Setting Realistic Milestones
Month 0-3: Probably no AI mentions
Month 4-6: Possible first mentions for specific queries
Month 7-9: Regular mentions for niche queries
Month 10-12: Consistent mentions for relevant queries
12+ months: Strong presence across relevant queries
Taking Action on Insights
Tracking is pointless without action. Monthly, review your data and ask:
If you're not being mentioned:
Review [why some brands show up in ChatGPT and others don't](/learn/why-brands-show-up-in-chatgpt) to diagnose the issue.
If you're mentioned inconsistently:
If you're mentioned with wrong information:
If competitors are mentioned more:
Study [the authority signals that make LLMs trust brands](/learn/authority-signals-llms-trust) to understand what to build.
The Bottom Line
You can't improve what you don't measure. Start with:
1. Manual testing of key queries (1 hour/week)
2. Google Alerts for mentions (5 min/day to review)
3. Simple spreadsheet dashboard (30 min/month to update)
4. Monthly review and strategy adjustment (1 hour/month)
That's less than 10 hours per month to stay informed about your AI visibility.
As automated tools become available, they'll make this easier. But starting with manual tracking teaches you what matters and gives you baseline data.
Keep learning: Now that you know how to track, follow our [step-by-step guide to getting your business into ChatGPT's knowledge base](/learn/step-by-step-chatgpt-visibility) to build what you'll be tracking. Understand [the authority signals LLMs trust](/learn/authority-signals-llms-trust) and [why some brands show up while others don't](/learn/why-brands-show-up-in-chatgpt).
New to AI search? Start with [how AI search differs from Google](/learn/how-ai-search-differs-from-google).
Ready to automate this? [Join our waitlist](/waitlist) to be notified when our LLM tracking tool launches. We'll handle the testing, tracking, and alerts so you can focus on building your business.