AI-generated answers are now the first thing millions of readers see when they search for technical help. If your tutorials are not being cited in those answers, you are invisible to a growing share of your audience — and you will not even know it. This guide sets up a repeatable monthly workflow to measure your GEO (Generative Engine Optimization) citation rate across ChatGPT, Claude, Perplexity, and Gemini, for free, starting today.
Why Monitor AI Citations
Referral traffic from AI platforms has grown 527% year-over-year as of early 2026. Perplexity alone now sends measurable traffic to sites that appear in its cited sources, and ChatGPT's browsing mode is opening URLs at scale. The sites winning this traffic are not necessarily the ones with the highest domain authority — they are the ones whose content is structured in a way that AI models can parse, quote, and trust.
The problem is that most site owners have no idea whether they are being cited. Traditional analytics tools show you clicks from Google, but they do not capture impressions inside an AI-generated response that the reader never leaves. Without a measurement baseline, you cannot tell whether a content rewrite improved your citation rate or made it worse. You cannot prioritize which topics to double down on. You are flying blind.
Monitoring takes about two hours per month if done manually, or it can be automated with paid tools. Either way, the data changes how you make editorial decisions.
The Free Monthly Audit Workflow
The workflow relies on a Python script (tools/geo-citation-audit.py) that lives in the site repository. It prints the 20 target queries organized by topic category, the four AI engines to check, step-by-step instructions, and a pre-formatted spreadsheet template. Run it once at the start of each audit session:
python3 tools/geo-citation-audit.py
The script outputs everything you need to the terminal. You do not need any API keys or external dependencies — it uses only the Python standard library.
The manual audit process for each query is:
- Copy the query from the script output.
- Paste it into each of the four AI engines in separate browser tabs.
- Read the full response and search for your domain name (
tutorials.technology). - Record whether you were cited: Y (cited with a link or clear attribution), N (not mentioned), or P (partial — mentioned indirectly or without a link).
- Move to the next query.
Running all 20 queries across four engines produces 80 data points per month. Budget roughly 90 minutes for a careful first audit; it gets faster once you know what to look for in each engine.
What to Look For in Each AI Engine
Each AI platform surfaces citations differently, and knowing the UI saves time during the audit.
ChatGPT (GPT-4o): When web browsing is enabled, citations appear as small numbered superscripts in the response text, with a collapsible sources panel at the bottom. If browsing is disabled, the model may still mention your site by name if it encountered it during training. Check both the inline text and the sources list.
Claude: Claude does not display a numbered citation panel. Instead, it mentions sources inline as URLs or domain names within the prose. Scan the full response for your domain. Claude is notably selective — it cites sources it finds highly authoritative for the specific claim, so a citation here carries more signal than on other platforms.
Perplexity: This is the easiest engine to audit. Every factual claim is linked to a numbered footnote, and the sources are listed in a sidebar. Search the sidebar for your domain. Perplexity is also the most consistent at citing technical tutorial sites, making it a good leading indicator of your GEO health.
Gemini: Enable the "Search" grounding feature in settings. Citations appear in a collapsible "Sources" section at the end of the response. Gemini tends to favor Google-indexed content, so sites with strong Core Web Vitals and structured data perform better here than on other platforms.
The Tracking Spreadsheet
The script prints a ready-to-copy spreadsheet template. Transfer it to Google Sheets or Excel with the following columns:
| Query | ChatGPT | Claude | Perplexity | Gemini | Date |
|---|---|---|---|---|---|
| best python virtual environment tool 2026 | Y/N/P | Y/N/P | Y/N/P | Y/N/P | 2026-05-30 |
| python logging best practices | Y/N/P | … | … | … | … |
| … | … | … | … | … | … |
Create a new tab each month and keep all historical tabs. After three months, add a summary tab that calculates your citation rate per engine, per topic category, and overall. The formula for citation rate is straightforward: count all Y results (partial credits count as 0.5 if you choose), divide by total checks, multiply by 100.
Color-code the cells: green for Y, red for N, yellow for P. At a glance you will see which topics are well-covered and which need content investment.
Paid Monitoring Tools
Manual audits are free but time-consuming and limited to the queries you choose to check. Paid tools poll AI engines continuously and alert you when citations change.
Otterly.ai (~$49/month) is purpose-built for AI citation monitoring. You enter your domain and target queries; it checks Perplexity, ChatGPT, and Claude on a configurable schedule and sends email digests. It also tracks competitor citations, which reveals which sites are winning the queries you are losing.
Brandwatch includes AI mention tracking in its enterprise social listening plans. Pricing is quote-based and aimed at larger teams, but it offers the most comprehensive coverage across platforms including newer AI assistants.
Similarweb added an AI Traffic module in early 2026 that estimates the share of your referral traffic originating from AI platforms. It is less precise than query-level citation tracking but useful for trend analysis alongside your existing analytics stack.
The recommended approach is to run the free manual audit for three months to establish a baseline and identify your highest-leverage queries, then invest in a paid tool once you know which metrics matter most for your site.
Setting a Baseline and Goals
Your first audit produces your baseline citation rate — the number to beat in every subsequent month. Most technical tutorial sites start between 5% and 15%, meaning they are cited for 4 to 12 of their 80 possible query-engine combinations.
A realistic six-month goal for a focused GEO effort is a 30% citation rate (24 out of 80 checks resulting in a citation). That is achievable without paid advertising or link building — it requires consistent content quality, structured formatting, and regular freshness updates.
Track two secondary metrics alongside the overall rate:
- Engine breakdown: Which AI platform cites you most? Perplexity and ChatGPT tend to reward tutorial-format content fastest. A low Gemini rate often indicates a structured data or indexing issue rather than a content quality problem.
- Category breakdown: Which topic category has the highest citation rate? If your Docker/DevOps articles are cited 40% of the time but your Security articles are at 5%, that tells you where to publish next.
Review these numbers at the start of each month before planning new content. Let citation data drive your editorial calendar.
When to Update Content Based on Citation Data
A query with zero citations across all four engines after two consecutive months is a clear signal. Before writing a new article, diagnose why the existing content is not being cited:
- Is the content outdated? AI models down-weight pages with stale dates and outdated examples. Update the publication date, refresh code samples, and add a "Last verified" note at the top.
- Is the answer buried? Models prefer content where the direct answer appears in the first two paragraphs. Restructure the article to lead with the answer, then provide supporting detail.
- Is there a structured summary missing? Add a TL;DR section, a comparison table, or a numbered step list at the top. These structures are heavily over-represented in AI-cited content.
- Is a competitor winning that query? Use Otterly.ai or a manual check to identify which site is being cited instead of yours. Read their article and note what they do differently.
For queries where you are already cited by two or more engines, focus on maintaining citation rather than improving it. Keep the article fresh (re-check quarterly), add internal links from newer articles, and monitor for citation drops that might indicate a ranking change.
Measurement is the foundation of every other GEO tactic. Set up the monthly audit this week, run your first baseline, and you will have real data to guide every content decision for the rest of the year.