Back to AI Visibility Tracking Hub
When a prospect asks ChatGPT “What are the best cybersecurity tools for mid-market companies?”, does your client’s brand appear in the answer? What about when they ask Perplexity for “top project management software for distributed teams”?
Most agencies have no visibility into these moments. They’re tracking Google rankings while buyers are getting recommendations from AI platforms that never show up in traditional analytics.
This guide provides a complete implementation framework for tracking brand mentions across AI search platforms. You’ll learn how to monitor visibility manually (the starting point for most agencies), when to graduate to automated tracking, and how to build a systematic brand monitoring workflow that reveals competitive gaps before clients even know to ask.
Why Does Brand Tracking in AI Search Matter?
Traditional brand monitoring tracks mentions across social media, news sites, and review platforms. AI search brand tracking adds a critical new layer: how your brand appears in AI-generated answers that shape purchase decisions.
The discovery shift:
73% of B2B buyers now begin product research with AI search, not Google. When someone asks “What CRM should I use for my e-commerce business?”, AI platforms synthesize answers from their training data and real-time web retrieval—either mentioning your brand or not.
Why it’s different from traditional monitoring:
| Traditional Brand Monitoring | AI Search Brand Tracking |
|---|---|
| Tracks mentions in published content | Tracks mentions in dynamically generated answers |
| Static mentions (once published, permanent) | Dynamic mentions (answers vary with each query) |
| Source attribution clear (article, tweet, review) | Source attribution complex (AI synthesizes multiple sources) |
| Sentiment analysis straightforward | Sentiment analysis requires context parsing |
| One-time capture | Requires repeated sampling for statistical validity |
Agency value: When you show clients where they’re visible—or invisible—in AI-generated answers competitors dominate, you’ve identified a competitive gap traditional SEO reports miss entirely.
What AI Platforms Should You Track?
Not all AI platforms have equal impact on buyer behavior. Prioritize based on adoption, citation behavior, and audience relevance.
Tier 1: Must-Track Platforms
1. ChatGPT (OpenAI)
- User base: 200M+ weekly active users
- Citation behavior: Rarely cites sources (only approximately 20% of mentions include links)
- Strength: Conversational, multi-turn research interactions
- Audience: General consumers, professionals, researchers
- Tracking priority: Critical
2. Perplexity
- User base: 15M+ monthly active users (growing rapidly)
- Citation behavior: Always cites sources (averages 5+ citations per answer)
- Strength: Real-time web search integration, transparent sourcing
- Audience: Researchers, professionals, technical users
- Tracking priority: Critical
3. Google AI Overviews (formerly SGE)
- User base: 13.14% of all Google searches now show AI Overviews
- Citation behavior: Blends brand mentions with source attribution
- Strength: Integrated into dominant search engine
- Audience: Broadest reach—all Google users
- Tracking priority: Critical
Tier 2: Important Secondary Platforms
4. Gemini (Google)
- User base: Integrated into Google ecosystem
- Citation behavior: Moderate citation rate, strong entity recognition
- Strength: Multimodal (text, image, video understanding)
- Tracking priority: High
5. Claude (Anthropic)
- User base: Growing among professionals and developers
- Citation behavior: High-quality synthesis, selective citations
- Strength: Long-context understanding, nuanced responses
- Tracking priority: Medium-High
6. Microsoft Copilot
- User base: Integrated into Microsoft 365, Bing
- Citation behavior: Bing-powered citations
- Strength: Enterprise integration
- Tracking priority: Medium (higher for B2B clients)
Tier 3: Emerging Platforms
- Grok (X/Twitter): Growing user base, X integration
- You.com: Privacy-focused search with AI features
- Brave Search AI: Privacy-centric alternative
Minimum viable tracking: ChatGPT + Perplexity + Google AI Overviews covers approximately 80% of B2B research behavior.
How Do You Track Brands Manually?
Manual tracking helps you understand AI response patterns and establish baselines before investing in automation. Most agencies start here.
Step 1: Build Your Query Library
Identify 15-25 conversational queries your target audience actually asks. These should span the buyer journey from awareness to decision.
Query sources:
- Customer support tickets (common questions)
- Sales call transcripts (prospect questions)
- Google Search Console “People Also Ask” data
- Reddit threads in your industry
- LinkedIn comment discussions
- Competitor comparison searches
Query structure:
- Awareness stage: “What is [category]?” / “How does [technology] work?”
- Consideration stage: “Best [category] tools for [use case]” / “Top [category] solutions compared”
- Decision stage: “[Your brand] vs [Competitor] which is better?” / “Is [Your brand] worth it?”
Example query library for project management software:
Awareness (5 queries):
- What is project management software?
- How does project management software help teams?
- What features should project management tools have?
- Why do companies use project management platforms?
- What are the types of project management tools?
Consideration (10 queries):
- Best project management software for remote teams
- Top project management tools for small businesses
- Project management software for marketing agencies
- Most popular project management platforms 2026
- Free project management tools with good features
- Project management software with time tracking
- Affordable project management solutions under $50/month
- Project management tools with Gantt charts
- Easy-to-use project management software for beginners
- Enterprise project management software compared
Decision (10 queries):
- Asana vs Monday.com which is better
- Is Asana worth the cost
- Trello vs Asana for marketing teams
- ClickUp vs Monday.com comparison
- Monday.com pricing vs Asana pricing
- Best alternative to Asana
- Asana reviews from actual users
- Why do companies switch from Trello to Asana
- Is Monday.com better than Asana for agencies
- Asana competitor comparison
Step 2: Run Queries Systematically
For each query, run it across your Tier 1 platforms and log results.
Manual tracking template (spreadsheet):
| Date | Platform | Query | Brand Mentioned? | Position | Citation Link | Competitors Mentioned | Sentiment | Notes |
|---|---|---|---|---|---|---|---|---|
| 2026-03-10 | ChatGPT | Best PM software for remote teams | Yes | 3rd mentioned | No | Monday, Trello, Asana, ClickUp | Neutral | Listed in top 5 |
| 2026-03-10 | Perplexity | Best PM software for remote teams | No | N/A | N/A | Asana, Monday, Trello | N/A | Not mentioned |
| 2026-03-10 | AI Overviews | Best PM software for remote teams | Yes | 2nd mentioned | Yes | Asana, [Brand], Monday | Positive | Cited for collaboration features |
Process:
- Open platform in private/incognito window (avoids personalization)
- Enter query exactly as written
- Review full AI response
- Log whether brand is mentioned
- Note position if listed (1st, 2nd, 5th, etc.)
- Record whether answer includes a citation link to your site
- List all competitor brands mentioned
- Assess sentiment (positive, neutral, negative)
- Add contextual notes
Time investment: Approximately 15-20 minutes per query (across 3 platforms). For 25 queries: 6-8 hours of manual work.
Step 3: Calculate Baseline Metrics
After running all queries, calculate core metrics:
1. Mention Rate
Mention Rate = (Queries where brand mentioned divided by Total queries) times 100
Example: Brand mentioned in 12 of 25 queries equals 48% mention rate
2. Citation Rate
Citation Rate = (Mentions with citation link divided by Total mentions) times 100
Example: 4 of 12 mentions included citation equals 33% citation rate
3. Share of Voice
Share of Voice = (Your brand mentions divided by Total brand mentions) times 100
Example: Your brand mentioned 12 times, competitors mentioned 38 times total equals 24% share of voice
4. Average Sentiment
Score each mention: Positive equals +1, Neutral equals 0, Negative equals -1. Calculate average.
Example: 5 positive, 6 neutral, 1 negative equals (5 minus 1) divided by 12 equals +0.33 average sentiment
Step 4: Identify Patterns
Look for patterns across your data:
Platform patterns:
- Strong on Perplexity (60% mention rate), weak on ChatGPT (20% mention rate) suggests good citation coverage but weak entity recognition
Query type patterns:
- High visibility for “What is [category]?” queries, low for “Best [category] for [use case]” suggests thought leadership presence but weak product positioning
Competitor patterns:
- Competitor A appears in 85% of answers, Competitor B in 67%, you in 48% indicates competitive gap analysis
Sentiment patterns:
- Positive mentions concentrated in “features” queries, neutral in “comparison” queries suggests strength in capabilities, neutral in head-to-head positioning
These patterns guide optimization priorities.
What Are the Limitations of Manual Tracking?
Manual tracking works for initial baselines but breaks down at scale.
Limitation 1: AI responses vary dramatically
Run the same query three times, you get three different answers. AI platforms use non-deterministic generation—responses change with each request. A single run has zero statistical validity.
Solution: Run each query 10-20 times to get reliable averages. (This is impractical manually.)
Limitation 2: Time investment scales poorly
25 queries times 3 platforms times 10 runs equals 750 manual checks. At 2 minutes per check equals 25 hours of work. Monthly tracking compounds this.
Solution: Automate repetitive queries.
Limitation 3: No historical trending
Manual spreadsheets track point-in-time snapshots but lack structured historical tracking. Hard to visualize trends over weeks/months.
Solution: Use dedicated tracking platforms.
Limitation 4: No alerts
If your mention rate drops 40% week-over-week, you won’t know until you manually re-run queries and compare.
Solution: Automated monitoring with threshold alerts.
How Do You Automate Brand Tracking?
Automated platforms run queries consistently, extract mentions systematically, and trend data over time.
What Automated Tools Track
Core capabilities:
- Scheduled query runs → 50-100 runs per query for statistical confidence
- Mention extraction → Automatically identifies brand mentions in responses
- Citation detection → Captures whether mentions include source links
- Competitor tracking → Identifies which competitors appear alongside you
- Sentiment analysis → Classifies mention tone
- Historical trending → Tracks changes week-over-week, month-over-month
- Alerts → Notifies when visibility drops or competitors surge
Leading AI Visibility Tracking Platforms
PhantomRank
- Strength: 45 strategic prompts across 9 intent types, Industry Metrics competitive scans
- Coverage: Deep Perplexity analysis, ChatGPT/Gemini/Grok on roadmap
- Pricing: Starts at $999/month
- Best for: Agencies managing multiple clients, competitive intelligence focus
SE Ranking
- Strength: Integrated AI visibility tracking with traditional SEO suite
- Coverage: Multiple AI platforms
- Best for: Agencies already using SE Ranking for traditional SEO
Ahrefs Brand Radar
- Strength: Brand mention tracking across web and AI platforms
- Coverage: Broad web monitoring + AI search layer
- Best for: Agencies needing combined brand monitoring
Siftly
- Strength: Conversational AI search focus
- Coverage: ChatGPT, Perplexity, Claude, others
- Best for: Agencies focused specifically on AI search visibility
When to Automate
Stick with manual tracking if:
- You’re still learning AI search patterns
- Client budget doesn’t support paid tools yet
- You track fewer than 10 clients
- Monthly spot-checks are sufficient
Graduate to automation when:
- You track 5+ clients regularly
- Clients want weekly/monthly AI visibility reports
- Manual tracking consumes 10+ hours per month
- You need historical trending data
- Competitive benchmarking is critical
How Do You Build a Brand Tracking Workflow?
Whether manual or automated, establish a systematic workflow.
Weekly Monitoring Workflow
For high-value clients:
Monday morning (30 minutes):
- Run top 10 most important queries (or review automated results)
- Check mention rates for each query
- Flag any major drops (over 20% decrease)
- Note new competitors appearing
Wednesday (15 minutes):
- Review sentiment scores
- Check for negative mentions
- Screenshot notable mentions for client reports
Friday afternoon (30 minutes):
- Calculate week-over-week changes
- Update tracking spreadsheet or dashboard
- Prepare client update if significant changes occurred
Monthly Reporting Workflow
Week 4 of each month (2-3 hours):
- Run full query library (25-50 queries)
- Calculate monthly metrics:
- Mention rate
- Citation rate
- Share of voice
- Sentiment score
- Compare to previous month:
- Mention rate change: +X% or -X%
- Share of voice change: +X points or -X points
- New competitors appearing
- Sentiment shifts
- Generate client report (see reporting section below)
- Recommend optimizations based on gaps
Quarterly Deep-Dive Workflow
Every 90 days (4-6 hours):
- Expand query library (add 10-15 new queries based on market changes)
- Competitive benchmark (run full analysis on top 3-5 competitors)
- Platform expansion (test new AI platforms like Grok, You.com)
- Optimization impact assessment (correlate content updates with visibility changes)
- Strategic planning (set visibility targets for next quarter)
How Do You Report Brand Tracking to Clients?
Frame brand tracking as competitive intelligence, not just metrics.
Executive Summary Template
AI Visibility Report: March 2026
Overall Performance:
- Mention Rate: 42% (up from 34% in February, +8 points)
- Citation Rate: 28% (up from 19%, +9 points)
- Share of Voice: 23% (up from 18%, +5 points)
Key Findings:
Strong momentum in product comparison queries. Your mention rate for “best [category] for [use case]” queries increased from 28% to 51%. Content optimization to comparison pages drove this improvement.
Competitive gap closing with Competitor A. Their share of voice decreased from 38% to 32% while yours increased from 18% to 23%. You’re now closer to them than to 4th place.
Low visibility in awareness-stage queries. Only 15% mention rate for “What is [category]?” queries. Opportunity to improve thought leadership presence.
Recommended Actions:
- Create comprehensive “What is [category]?” guide with extractable facts
- Add structured data to comparison pages (boosting citation rate)
- Continue optimizing product positioning content (sustain momentum)
Detailed Metrics Dashboard
Create a simple dashboard visual:
Mention Rate Trend (Last 6 Months)
50% | ●
45% | ● /
40% | ● / /
35% | ● / / /
30% | ● / / / /
25% | ● / / / / /
|------|-----|-----|-----|-----|
Oct Nov Dec Jan Feb Mar
Share of Voice by Competitor
Competitor A: ████████████████████████████ 32%
Your Brand: ███████████████ 23%
Competitor B: ██████████████ 21%
Competitor C: ████████ 14%
Others: ██████ 10%
Platform-Specific Insights
Break down performance by platform:
| Platform | Mention Rate | Citation Rate | vs. Last Month |
|---|---|---|---|
| ChatGPT | 38% | 15% | +12% ↑ |
| Perplexity | 51% | 89% | +6% ↑ |
| AI Overviews | 35% | 42% | -3% ↓ |
Insight: Strong on Perplexity (high citation coverage), weaker on ChatGPT (entity recognition opportunity).
What Are Common Tracking Mistakes?
Avoid these pitfalls that undermine tracking validity.
Mistake 1: Single-Run Queries
The error: Running each query once and treating results as definitive.
Why it’s wrong: AI responses vary dramatically run-to-run. Single runs have no statistical validity.
Fix: Run each query 10+ times (manually) or 50+ times (automated) to get reliable averages.
Mistake 2: Personalized Results
The error: Running queries while logged into personal accounts, using regular browser windows.
Why it’s wrong: AI platforms personalize responses based on browsing history and account data. Your results won’t match what prospects see.
Fix: Always use private/incognito mode, log out of accounts, or use automation that randomizes user agents.
Mistake 3: Ignoring Competitors
The error: Only tracking whether your brand appears, not who appears alongside you.
Why it’s wrong: Mention rate without competitive context is meaningless. 40% mention rate is strong if competitors are at 20%, weak if they’re at 70%.
Fix: Always log competitor mentions, calculate share of voice.
Mistake 4: No Historical Baseline
The error: Starting tracking without establishing a baseline period.
Why it’s wrong: You can’t measure improvement without a starting point. “43% mention rate” means nothing without context.
Fix: Run 60-90 days of baseline tracking before making optimization claims.
Mistake 5: Static Query Lists
The error: Using the same 25 queries for 6+ months without updating.
Why it’s wrong: Market language evolves, new competitors emerge, buyer questions shift.
Fix: Refresh query library quarterly. Add 10-15 new queries, retire outdated ones.
How Does Brand Tracking Connect to Optimization?
Brand tracking identifies gaps. Generative Engine Optimization (GEO) fills them.
Workflow:
- Track → Measure current visibility, identify competitive gaps
- Analyze → Determine why competitors get mentioned and you don’t
- Optimize → Implement content quality improvements, on-page optimization, and technical fixes
- Re-track → Measure impact, iterate
Example optimization cycle:
- Week 0: Baseline tracking shows 28% mention rate
- Week 1-2: Analyze why competitors get cited (better comparison tables, more specific statistics)
- Week 3-6: Optimize top 10 pages (add tables, front-load facts, improve extractability)
- Week 7-8: Re-track shows 39% mention rate (+11 points)
- Week 9+: Repeat cycle on next priority pages
Tracking without optimization is just reporting. Optimization without tracking is guesswork.
What Results Should You Expect?
Brand tracking reveals current state. Optimization drives improvement.
Realistic improvement timeline:
Month 1: Establish baseline, no change expected Month 2: Begin optimization, minimal visibility change (content indexing lag) Month 3: First improvements visible (10-15% mention rate increase typical) Month 4-6: Sustained improvement (20-30% mention rate increase, 10-15 point share of voice gain) Month 7+: Compounding gains as more content optimized
PhantomRank customers tracking systematically see:
- 20-35% mention rate improvement within 90 days
- 15-25% citation rate improvement within 90 days
- 10-15 point share of voice gain within 6 months
These gains require consistent optimization, not just tracking.
What’s Next: From Tracking to Strategy?
Once you’ve established baseline tracking and identified competitive gaps, the next step is building a complete AI visibility tracking framework that connects measurement to optimization to reporting.
For agencies looking to integrate this into client services systematically, see our guide on how agencies can sell AI visibility tracking services.
Ready to see where your clients stand in AI search before their competitors do?
PhantomRank’s AI Visibility Tracker monitors 45 strategic prompts across 9 intent types, giving agencies a complete competitive intelligence picture in minutes—not hours of manual work.
Get Access or See How It Works.
Related Resources:
- The Complete Guide to AI Visibility Tracking — Full measurement framework
- AI Search vs Traditional Search — Understanding the differences
- Content Quality Checker for AI Citations — Optimize content for citations
- What Is Generative Engine Optimization? — Strategic optimization framework
Articles in This Topic
5 articles exploring this topic in depth.
Citation Velocity: The Leading Indicator Your Agency Isn't Tracking Yet
Discover why Citation Velocity is the ultimate predictive metric for AI visibility, and how tracking it can protect your clients from emerging competitors.
Conversational Retention: The AI Visibility Metric Your Dashboard Is Missing
Most AI tracking tools measure single-prompt responses. Learn what Conversational Retention is, why it matters more, and how to start measuring it for your clients.
Entry-Point Dominance: Why ToFu AI Prompts Are Your Most Valuable Real Estate
Winning a Category Awareness prompt in AI search is mathematically worth more than any downstream prompt. Here's the case for tracking entry-point dominance.
How to Calculate Share of Synthesis for Your Clients
Move beyond simple brand mentions. Learn how to calculate Share of Synthesis and prove the ROI of being the primary cited source in AI search.
Why the First Brand Mentioned in an AI Chat Session Wins the Sale
The brand named first in an AI conversation anchors the entire session. Learn why entry-point dominance in multi-turn AI chat is your highest-value real estate.