How to Track AI Visibility Without Paying for Tools

The three questions hiding inside "AI visibility"

Before you track anything, it helps to know what you are actually trying to measure. Most guides treat AI visibility as one thing. It is three.

Am I mentioned? When someone asks ChatGPT or Perplexity a question in your category, does your brand name appear in the answer? This is the simplest check and the one most people start with.

Am I cited with a link? A mention without a link is a name-drop. A citation with a live URL is a traffic source. Perplexity almost always shows source links. ChatGPT rarely does unless it used browsing mode. The distinction matters because a Superlines analysis found that 73% of AI brand presence consists of "ghost citations" where AI links to your site without ever naming your brand. If you only check for your brand name in the text, you miss most of your actual visibility.

What is my share of voice? Of all the brands mentioned in answers to your category questions, what percentage are yours? This is the metric that tells you whether you are gaining or losing ground against competitors over time.

Each question needs a different tracking method. Mentions you can check manually. Citations require checking the source links in each answer. Share of voice requires logging every competitor mentioned across every prompt, every month, and doing the math. This guide covers all three.

The stakes are worth the effort. Conductor's 2026 data shows AI search visitors convert at twice the rate of normal organic visitors. Resoneo's study of 27,000 ChatGPT responses found the March 2026 model switch dropped unique domains cited per answer from 19.1 to 15.2. Fewer slots. Higher conversion rates. The brands that show up in those answers have a measurable revenue advantage over the ones that do not.

Step 1. Set up the free dashboards you already have access to

Before you open ChatGPT and start typing prompts, set up the three free data sources that track AI visibility passively. These run in the background and collect data without you doing anything manually each week.

Google Search Console generative AI report

On June 3, 2026, Google launched a dedicated generative AI performance report inside Search Console. This is the single biggest free AI visibility tool available right now.

The report shows how often your URLs appeared inside AI Overviews, AI Mode, and generative AI features in Discover. It breaks the data down by page, country, device, and date with hourly, daily, weekly, and monthly granularity. Data starts from May 18, 2026.

What it does not show (yet) is click data, CTR, average position, or query-level breakdowns. Google has said it will add more metrics over time, but for now you get impressions only. That is still useful. It tells you which of your pages Google's AI is pulling from, whether that number is growing or shrinking, and which countries see your content in AI answers.

To find it, open Search Console, go to Performance, and look for "Generative AI" underneath the standard "Search results" report. If you do not see it yet, the rollout started with UK sites and is expanding globally. Check back every few weeks.

One important setting to check while you are there. Under Settings, look for AI controls and then Search generative AI. The default is "include," meaning your content can appear in AI answers. Leave it on. Opting out removes your content from AI Overviews and AI Mode entirely, and you lose both the impressions and any traffic that comes with them. Opting out does not affect your standard organic rankings.

Bing Webmaster Tools AI performance

Bing Webmaster Tools launched its own AI Performance report in beta on February 10, 2026. It actually shows more granular data than Google's version.

Bing's report includes direct citations (how many times your site was explicitly cited in a Copilot AI answer) and grounding queries (the specific queries that triggered your content to be used as grounding for an AI answer). This is closer to real citation tracking than Google's impression-only data.

Because Copilot powers search across Windows, Microsoft 365, and Edge, these metrics represent your visibility in the most widely deployed enterprise AI assistant. If your audience skews toward office workers and B2B buyers, Bing's data may matter more than you expect.

Use both reports together. Google gives you volume (AI Overviews reach over 2.5 billion monthly active users). Bing gives you citation-level detail. Neither gives you the full picture alone.

GA4 referrer filtering for AI traffic

Your Google Analytics 4 account already collects referrer data from AI engines. The problem is that it buries AI traffic inside "direct" or "unassigned" because most AI engines strip or modify referrer headers. But some traffic does come through with identifiable referrers, and you can filter for it.

Add these domains as custom channel groupings or filtered views in GA4 to start isolating whatever AI referral traffic is identifiable.

The volume will start small. Track the trend over months, not the absolute number. A growing AI referral line in your analytics is a leading indicator that your content is being cited more frequently, even before your manual prompt testing confirms it.

Important caveat. Most AI-referred visits still show as "direct" because the engines strip headers. GA4 referrer filtering catches maybe 10-30% of actual AI traffic depending on the engine and how the user clicked through. It is a floor, not a ceiling. For full AI referral attribution on WordPress, that is what CiteTrack's revenue attribution does with signals beyond the referrer header.

Step 2. Build a prompt bank that mirrors real questions

The three dashboards above track AI visibility passively. The prompt bank tracks it actively. This is where you manually ask AI engines the questions your customers ask and log whether you show up.

Split your prompt bank into four buckets: direct brand prompts ("What is [your brand]?"), category prompts ("Best [category] for [audience]"), problem-first prompts ("How do I fix [problem]"), and comparison prompts ("[Brand] vs [competitor]"). Aim for 20 to 30 prompts total. Lock the wording and reuse it every time you test.

Write every prompt the way an actual person would type it. "Best CRM for a 5-person agency" beats "CRM software recommendations for small businesses."

Step 3. Set up a free logging spreadsheet

One spreadsheet with the right columns, set up once: Date tested, Prompt used, Engine, Brand mentioned (Y/N), Link to your site (Y/N), Position, Competitors mentioned, Source URLs cited, Sentiment, and Notes. Make one row per prompt per engine per test date. The "Link to your site" column catches ghost citations that brand-name-only scanning misses.

Step 4. Run the tests on a real schedule

Weekly: top 5 prompts across 4 engines (20 checks, 30-40 minutes). Monthly: full 20-30 prompt bank (2-3 hours). Always test logged out or incognito. ChatGPT has 55-60% of AI referral traffic. Perplexity always shows source citations. Start with those two, add Gemini and Claude as cadence allows.

Step 5. Track competitors and build share of voice by hand

Pick 3-4 real competitors. Log every competitor mentioned in every answer. Your share of voice = your mentions / total mentions. 12 out of 80 = 15%. Directionally accurate and free.

Step 6. Use free visibility checkers to fill the gaps

Ahrefs AI Visibility Checker (5 free checks/day), Semrush AI Visibility Checker (free, no signup), HubSpot AEO Grader (free), SE Ranking (5 checks/day), and CiteTrack's 10 free tools + Chrome extension. The first four check whether AI mentions your brand. CiteTrack's tools check whether your pages are structurally ready to be cited.

Step 7. Turn what you find into actual content fixes

Missing from category prompts? Build a direct answer page. Competitors listed first? Check their answer-first structure. Winning brand but losing comparisons? Build a "[you] vs [competitor]" page. Notes column flagging outdated info? Update the page, do not rebuild it. The CiteTrack Chrome extension generates a prioritized fix list for any page.

The missing middle: automating before paying for a tool

n8n or Make.com can automate prompt testing via API at pennies per prompt. Not truly free, but dollars per month instead of hundreds. When that maintenance cost exceeds what a tool saves, CiteTrack at $59/yr automates the full workflow inside WordPress.

Where manual tracking breaks down

1 brand, 25 prompts, monthly = 3 hours (manual wins). Weekly = 8-10 hours ($59/yr saves ~$2,400/yr in time). Multiple brands = paying is arithmetic. Average tool cost in 2026 is $337/month. CiteTrack at $59/yr is the cheapest path from manual to automated for WordPress.

Frequently asked questions

Where to go from here

The Crawl Academy explainer covers GPTBot, ClaudeBot, and PerplexityBot. The AEO guide and GEO guide cover the fix-and-measure loop. The citation tracker comparison covers every paid option.

Guides · how to track ai visibility without paying for tools