Measuring AEO: How to Track Citations in ChatGPT and Perplexity

The Problem: You're Invisible to AI Search Engines

You shipped. Your product works. Your organic traffic from Google is growing. But you're not in ChatGPT's answers. You're not cited by Perplexity. And your competitors are.

Here's the brutal truth: traditional SEO tools like Ahrefs, Semrush, and Surfer SEO don't measure what matters anymore. They track Google rankings, backlinks, and keyword positions. None of that tells you if Claude is citing your content, if ChatGPT is recommending your product, or if Perplexity users are seeing your brand as a source.

Answer Engine Optimization (AEO) is different. It's not about ranking higher on Google—it's about being the source AI systems trust enough to cite. And if you can't measure it, you can't improve it.

This guide walks you through a founder-friendly tracking system for AI citations. No agency needed. No expensive tools. Just a repeatable process you can run weekly to know exactly where you stand across ChatGPT, Perplexity, Claude, and Gemini.

Why Traditional SEO Tools Miss AEO Entirely

Ahrefs and Semrush built their empires on Google's API. They crawl the web, track rankings, measure backlinks, and give you a dashboard. That model works for traditional search. It doesn't work for AI.

AI systems don't publish a ranking algorithm. They don't have a public API for citations. ChatGPT doesn't tell you when it's citing your site. Perplexity doesn't rank you in a leaderboard. Claude doesn't show you citation metrics in a dashboard.

Instead, AI citations happen inside a black box. The only way to measure them is to ask the AI yourself, track the responses, and aggregate the data over time.

Traditional SEO tools tried to adapt. They added "AI search" features. But they're still measuring Google fundamentals—domain authority, backlink quality, keyword density. None of that correlates with AI citations at the rate it used to correlate with Google rankings.

You need a different approach. One built for AI.

Prerequisites: What You Need Before You Start

Before you build your AEO tracking system, you need three things:

1. Access to the AI Systems You Care About

You need active accounts with:

ChatGPT (free or Plus; Plus gives you web browsing and better citation rates)
Perplexity (free or Pro; Pro includes more sources and better citations)
Claude (free via Claude.ai or through Anthropic's API)
Google Gemini (free; optional but recommended)

If you're bootstrapped, start with ChatGPT and Perplexity. Those two account for roughly 70% of AI search traffic among founders and technical users.

2. A Spreadsheet or Database to Track Results

You can use Google Sheets, Notion, Airtable, or a simple CSV file. You'll need columns for:

Query (the search term)
AI System (ChatGPT, Perplexity, Claude, Gemini)
Date
Cited (Yes/No)
Citation Position (1st, 2nd, 3rd, etc.)
Exact Citation Text
URL Cited
Notes

Keep it simple. You'll update this weekly.

3. A List of Core Queries Your Business Cares About

These are the questions your customers ask before they find you. For a SaaS product, these might be:

"Best [product category] for [use case]"
"How to [your main value prop]"
"[Your product] vs [competitor]"
"Alternatives to [competitor]"
"How to [pain point you solve]"

Start with 15-25 queries. You'll test these weekly. More on this below.

Step 1: Define Your Core Keyword Set for AEO Tracking

This is where most founders mess up. They track too many keywords, or they track the wrong ones.

AEO citations happen for different queries than Google rankings. A query like "best project management software" might rank you on Google but never get you cited in ChatGPT. Meanwhile, "Asana alternatives" might cite you in Perplexity but not show you in Google's top 10.

Start narrow. Focus on queries where:

Your product is a direct answer. If you sell a password manager, track "best password managers," "password manager comparison," and "1Password alternatives." Don't track "cybersecurity best practices"—you're not the answer to that.

The query has commercial intent. Informational queries ("how does encryption work") rarely cite products. Comparison queries ("Notion vs Obsidian") almost always do.

You have content about the topic. If you haven't written about it, you won't get cited for it. This is where SEOABLE's AI blog generation becomes useful—you can generate 100 posts covering your core queries in under 60 seconds, then track citations for each.

Build your core set like this:

List 5-10 direct competitors
For each competitor, add "[Competitor] alternatives"
Add your product category + "best," "comparison," "vs"
Add your main use cases + "how to"
Add pain points you solve + "solution"

Example for a CRM tool:

"Best CRM for startups"
"Salesforce alternatives"
"HubSpot vs Pipedrive"
"How to choose a CRM"
"CRM for sales teams"
"Affordable CRM software"
"HubSpot alternatives"
"Pipedrive alternatives"
"Zoho CRM alternatives"
"Best CRM for small business"

That's 10 queries. Start here. Track these weekly. Add more once you have a baseline.

Step 2: Set Up Your Weekly Citation Audit

This is the operational core of your AEO measurement system. You'll do this every week, same day, same time. It takes 45 minutes to an hour for 15-20 queries across 2-3 AI systems.

The Process:

Open ChatGPT first. Use the Plus version if you can afford it ($20/month). The free version works, but Plus has better web access and higher citation rates. Log in. Make sure "Web Browsing" is enabled in settings.

Search your first query. Type it exactly as you would if you were a customer. "Best project management software for remote teams." Don't say "cite sources." Don't ask for citations. Just ask the question naturally.

Read the response carefully. Look for your domain name. Look for your product name. Look for your brand. ChatGPT will cite sources at the bottom of the response, often with links. Sometimes citations are inline in the text.

If you're cited:

Note the position (1st source, 2nd, 3rd, etc.)
Copy the exact text where you're cited
Note the URL it cited
Mark "Yes" in your spreadsheet

If you're not cited:

Check if competitors are
Note who is cited instead
Mark "No" in your spreadsheet
Save the full response (screenshot or copy-paste) for later analysis

Repeat for Perplexity. Go to Perplexity.ai. Use Pro if possible (free tier works but has rate limits). Ask the same query. Follow the same tracking process. Perplexity citations are usually clearer—they show source cards with URLs and publication names.

Optional: Repeat for Claude and Gemini. If you want comprehensive data, do the same for Claude and Gemini. You can use Claude.ai for free, or the Claude API if you're running this at scale. Gemini is free at Google Gemini.

Pro Tip: Batch Your Queries

Don't test one query across all four systems, then move to the next. Instead, test all your queries in ChatGPT, then move to Perplexity, then Claude. Your brain will adapt to each system's response format, and you'll spot citations faster.

Pro Tip: Use Incognito Mode

AI systems may personalize responses based on your account history. Use a private/incognito window to simulate a fresh user. This gives you cleaner data.

Pro Tip: Test at Different Times

AI systems sometimes pull different sources at different times of day. If you want rigorous data, test each query twice—once in the morning, once in the evening. Compare results. This matters more for Perplexity (which pulls live search results) than ChatGPT (which uses training data).

Step 3: Categorize and Analyze Your Citation Data

After your first week, you'll have raw data. Now you need to turn it into insights.

Calculate Your Citation Rate by Query:

For each query, count how many AI systems cited you:

Cited by 0/4 systems = 0% citation rate
Cited by 1/4 systems = 25% citation rate
Cited by 2/4 systems = 50% citation rate
Cited by 4/4 systems = 100% citation rate

Track this by week. Your goal is to increase this number over time.

Track Citation Position:

Being the 1st source cited is better than being 3rd. Create a separate metric: "Average Citation Position." If you're cited 1st, 2nd, and 3rd across different queries, your average is 2nd. Track this weekly. It should trend downward (toward 1st).

Identify Your Gaps:

Look at queries where you're not cited. Ask: Why? Common reasons:

You don't have content about this topic
Your content exists but isn't optimized for AI citations
Competitors have better, more recent content
Your site structure or schema markup isn't AI-friendly

This tells you what to write next. This is where SEOABLE's keyword roadmap becomes valuable—it prioritizes which content gaps matter most for your business.

Benchmark Against Competitors:

When you see citations, note who's being cited instead of you. Build a competitor citation tracker:

| Query | Your Citations | Competitor A | Competitor B | Competitor C | |-------|----------------|--------------|--------------|---------------| | Best CRM | 1/4 | 3/4 | 4/4 | 2/4 | | HubSpot alternatives | 0/4 | 3/4 | 2/4 | 1/4 |

This shows you exactly where you're losing ground. If a competitor is cited 4/4 times and you're cited 0/4, that's a content and optimization gap worth fixing.

Step 4: Understand Why Perplexity and ChatGPT Cite Differently

This is critical. ChatGPT and Perplexity have fundamentally different architectures. Understanding this changes how you optimize.

ChatGPT's Citation Model:

ChatGPT uses a training dataset (knowledge cutoff in April 2024 as of this writing). When you ask a question, ChatGPT generates an answer based on patterns in that training data. Then, if you have web browsing enabled, it can search the web to verify or update its answer.

Citations in ChatGPT happen when:

You ask for sources explicitly ("cite your sources")
ChatGPT uses web browsing to verify a claim
The answer requires recent information

ChatGPT prioritizes:

Authority and domain reputation
Freshness (if it did a web search)
Relevance to the exact query

Perplexity's Citation Model:

Perplexity is built on real-time web search. Every answer includes a web search. Perplexity's algorithm looks at search results, ranks them by relevance, and cites the top sources.

Citations in Perplexity happen when:

Your page ranks in the top results for the query
Your page has relevant content
Your schema markup is correct

Perplexity prioritizes:

Search ranking (like Google)
Schema markup and structured data
Content freshness
Page speed and technical SEO

The key insight: Perplexity's citation model is closer to Google than ChatGPT's. If you rank well in Google for a query, you'll likely be cited by Perplexity. But ChatGPT citations require different optimization.

This is why your alternatives page is your highest-converting asset—it's the one page type that gets cited consistently across both systems.

Step 5: Optimize Content for AEO Citations

Once you know where you're not being cited, you need to fix it. Here's what actually works:

Answer-First Content Structure:

AI systems cite sources that directly answer the question. If your article buries the answer in a 2,000-word essay, you won't get cited.

Instead, structure like this:

Direct answer (1-2 sentences)
Why this answer (2-3 sentences)
Supporting details
Comparison or alternatives

Example: If the query is "best CRM for startups," your first paragraph should be: "[Your Product] is the best CRM for startups because it combines affordability, ease of use, and sales automation in a single platform. Most startups save $500/month compared to Salesforce while cutting implementation time by 60%." Then expand.

AI systems cite this immediately. Buried answers get skipped.

Schema Markup for AI Citations:

Structured data directly impacts AI citation rates. Perplexity especially relies on schema to understand your content.

Add this schema to your pages:

Product schema (for product pages):

&#123;
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Your Product",
  "description": "Short description",
  "brand": &#123;
    "@type": "Brand",
    "name": "Your Brand"
  &#125;,
  "offers": &#123;
    "@type": "Offer",
    "price": "99",
    "priceCurrency": "USD"
  &#125;
&#125;

Article schema (for blog posts):

&#123;
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "Your Article Title",
  "author": &#123;
    "@type": "Person",
    "name": "Author Name"
  &#125;,
  "datePublished": "2024-01-15",
  "dateModified": "2024-01-20"
&#125;

FAQPage schema (for comparison pages):

&#123;
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    &#123;
      "@type": "Question",
      "name": "What is [Product]?",
      "acceptedAnswer": &#123;
        "@type": "Answer",
        "text": "[Your answer]"
      &#125;
    &#125;
  ]
&#125;

Add this schema to your pages. Perplexity will cite you more. ChatGPT will trust your content more.

Recency and Updates:

AI systems favor fresh content. If your article is from 2022 and a competitor's is from 2024, the competitor wins.

Update your top pages every 3-6 months. Change the date. Add new data. Refresh the examples. This signals freshness to both AI systems and Google.

Content Patterns That Earn Citations:

Research shows certain content patterns consistently earn AI citations:

Comparison tables – AI systems cite pages with clear side-by-side comparisons
"X Alternatives" pages – Perplexity and ChatGPT cite these almost 100% of the time
"Best of" roundups – "Best CRM for startups" gets cited more than "CRM features explained"
How-to guides with steps – Numbered steps get cited more than narrative guides
Pricing comparison – If you include pricing, you get cited more
Use case-specific content – "CRM for sales teams" gets cited more than "What is a CRM?"

Focus on these patterns. They work.

Step 6: Build a Long-Term AEO Dashboard

After 4-8 weeks of weekly tracking, you'll have enough data to build a real dashboard. Here's what to track:

Weekly Metrics:

Citation rate by query (% of AI systems citing you)
Average citation position (trending toward 1st)
New citations (queries where you weren't cited last week, now you are)
Lost citations (queries where you were cited, now you're not)

Monthly Metrics:

Total citations across all queries and systems
Citation rate trend (should be increasing)
Competitor citation rates (to benchmark)
Content gaps (queries you're not cited for)

Quarterly Metrics:

Which content updates moved the needle
Which AI systems cite you most (Perplexity vs ChatGPT)
Which query categories perform best (alternatives vs how-to)
ROI on content investment (citations per post)

Use a simple Google Sheet or Notion database. Update it weekly. Review it monthly. This becomes your AEO operating system.

Step 7: Connect AEO Citations to Business Outcomes

Here's the question nobody asks: Does getting cited by ChatGPT actually drive revenue?

Yes. But you need to measure it.

Set up UTM parameters for AI traffic. When you get cited, the AI system includes a link. Visitors click that link and come to your site. You need to track this.

Add UTM parameters to your links:

If you're on Perplexity, you don't control the link. But you can see traffic coming from Perplexity in your analytics. Set up a filter:

Source: perplexity.ai
Medium: referral

If you're in ChatGPT, same thing:

Source: openai.com (or chat.openai.com)
Medium: referral

Track the funnel:

Citations (measured weekly via your audit)
Clicks (measured via analytics)
Signups (measured via conversion tracking)
Revenue (measured via payment data)

Correlate these. If your citations increase by 50%, do your signups increase? By how much?

One founder we tracked (Solo Founder Hits 50K Organic/mo in Four Months) found that AEO citations drove 30% of their organic traffic after 4 months. That's real impact.

Advanced: Multi-LLM Citation Tracking

Once you master ChatGPT and Perplexity, expand to Claude and Gemini. The process is the same, but the nuances are different.

Claude (via Claude.ai):

Claude doesn't show explicit citations like ChatGPT
It references sources inline in the text
You have to read carefully to find mentions of your brand or domain
Claude is more verbose; look for your brand name in longer responses
Claude is less likely to cite products; more likely to cite educational content

Google Gemini:

Gemini shows citations but they're sometimes unclear
Gemini pulls from Google Search, so Google rankings matter
If you rank in Google's top 5, you'll likely be cited by Gemini
Gemini is newer; citation patterns are still evolving

The Comprehensive Tracking Spreadsheet:

Once you track all four systems, your spreadsheet might look like:

| Query | Date | ChatGPT | Perplexity | Claude | Gemini | Avg Position | Trend | |-------|------|---------|------------|--------|--------|--------------|-------| | Best CRM | Week 1 | Yes (2nd) | Yes (1st) | No | Yes (3rd) | 2nd | ↑ | | Best CRM | Week 2 | Yes (1st) | Yes (1st) | No | Yes (2nd) | 1.3rd | ↑ | | HubSpot alt | Week 1 | No | Yes (2nd) | No | No | 2nd | → | | HubSpot alt | Week 2 | Yes (3rd) | Yes (1st) | No | No | 2nd | ↑ |

This gives you a complete picture. You can see which systems cite you, which don't, and whether you're improving.

Pro Tips: Avoid These Common Mistakes

Mistake 1: Testing Only Once

AI responses vary. Sometimes you're cited, sometimes you're not, even for the same query. Test each query at least twice per week to reduce noise.

Mistake 2: Not Updating Your Content

Getting cited once doesn't mean you'll keep getting cited. Update your top pages every quarter. Add new data. Refresh examples. This keeps you relevant.

Mistake 3: Ignoring Perplexity

Founders focus on ChatGPT because it's more famous. But Perplexity citations are easier to track (they're explicit) and easier to earn (they follow Google rankings). Start with Perplexity. It's your faster path to AEO success.

Mistake 4: Forgetting About Schema

Schema markup seems technical and boring. It's not. Structured data directly impacts AI citation rates. Add it. It takes 30 minutes. It compounds over time.

Mistake 5: Chasing Every Query

You can't rank for everything. Focus on 15-25 core queries. Master those. Then expand. Spreading yourself thin across 100 queries wastes time.

Connecting AEO Measurement to Your Overall SEO Strategy

AEO isn't separate from SEO. It's the next layer.

Traditional SEO (Google rankings) is still important. But traditional SEO tools miss AEO entirely. You need both.

Here's how they fit together:

Google Search (Traditional SEO):

Measures: Rankings, traffic, conversions
Tools: Google Analytics, Google Search Console, Ahrefs, Semrush
Timeframe: 3-6 months to see results
Value: Still drives 70% of organic traffic for most startups

AI Search (AEO):

Measures: Citations, positions, presence in answers
Tools: Manual tracking (this guide) or emerging AEO platforms
Timeframe: 4-8 weeks to see improvements
Value: Growing fast; will be 20-30% of organic traffic within 2 years

Don't abandon Google SEO. Do both. Use this guide to measure AEO. Use SEOABLE's SEO audit to measure Google SEO. Together, they give you complete visibility.

The Seoable Advantage: Automation at Scale

Manual citation tracking works. It's rigorous. It teaches you how AEO works.

But it doesn't scale. If you have 50 core queries, testing them twice weekly across 4 AI systems is 400 manual tests per month. That's 10+ hours of work.

This is where SEOABLE's AI Engine Optimization platform changes the game. You enter your domain. In under 60 seconds, you get:

A complete domain audit (technical SEO, content gaps, competitive analysis)
A brand positioning report (where you stand vs competitors)
A keyword roadmap (which queries to target, in priority order)
100 AI-generated blog posts (covering your core keywords)

The 100 posts are built for both Google and AEO. They follow the answer-first structure. They include schema markup. They're designed to get cited.

Then, you layer your manual AEO tracking on top. You know exactly which posts are driving citations. You can measure ROI per post. You can iterate faster.

For founders who ship, this is the fastest path from zero to AEO visibility. One payment. One hour of setup. Then you measure weekly and optimize.

Summary: Your AEO Measurement Checklist

Week 1: Setup

Define your 15-25 core queries
Create your tracking spreadsheet
Set up accounts with ChatGPT Plus and Perplexity Pro
Run your first weekly audit (all queries, 2-3 systems)

Weeks 2-4: Establish Baseline

Run weekly audits (same day, same time)
Track citation rates and positions
Identify your top-cited queries and your gaps
Note which competitors are cited instead of you

Weeks 5-8: Optimize

Update your top pages with answer-first structure
Add schema markup to key pages
Create new content for high-gap queries
Retest and track improvements

Month 3+: Scale and Measure

Build a dashboard (weekly + monthly metrics)
Expand to Claude and Gemini
Connect citations to business outcomes (traffic, signups, revenue)
Double down on what works

Key Takeaways:

AEO is measurable. You don't need expensive tools. Manual tracking works and teaches you how AI citation works.
Perplexity is easier than ChatGPT. Start there. Perplexity citations follow search rankings. ChatGPT citations require different optimization.
Answer-first content wins. Structure your pages to answer the question immediately. AI systems cite these.
Schema markup matters. Add structured data. It's technical but it works.
Weekly tracking compounds. Four weeks of data is interesting. Eight weeks is actionable. Twelve weeks shows trends.
Citations drive traffic. Measure the funnel: citations → clicks → signups → revenue. AEO is only valuable if it converts.
Content gaps are opportunities. If competitors are cited and you're not, that's a content gap. Fix it. Track the improvement.

Start this week. Pick your 15 core queries. Run your first audit. Track one week of data. You'll see patterns immediately. Then iterate.

The founders winning at AEO right now aren't the ones with the biggest budgets. They're the ones measuring weekly and optimizing based on data. You can do the same. No agency needed. No expensive tools. Just rigor and repetition.

Ship your AEO tracking system. Then ship better content. Then measure again. That's how you become visible to AI search engines.