Measuring AEO Performance

    How to track whether your content is getting cited by ChatGPT, Claude, and Perplexity — methods, what to measure, and expected timelines.

    Publishing AEO-optimized content is step one. Measuring whether it is working is step two.

    Method 1 — Direct query testing

    Go to ChatGPT, Claude, and Perplexity. Ask the exact question your article answers. Check whether your domain or article is cited in the response. Run this test weekly for your 5 most important target queries and track in a spreadsheet:

    QueryChatGPTClaudePerplexityDate
    What is AEOYesNoYes2026-03-08

    Method 2 — Analytics referrer tracking

    • Perplexity passes referrer data to your analytics. In GA4, check Traffic Acquisition for sessions with source containing 'perplexity.ai'.
    • ChatGPT Browse appears as source 'chatgpt.com' in GA4. Set up a custom channel grouping to capture all AI-referred traffic separately from organic and direct.

    Method 3 — GPTBot crawl verification

    In your server access logs, filter for User-Agent containing 'GPTBot'. Regular GPTBot visits to your article URLs confirm your content is being indexed for ChatGPT. If GPTBot has never visited a page, that page cannot be cited.

    Expected timeline — new domain with 6–10 articles

    PeriodWhat happens
    Week 1–2GPTBot begins crawling — verify in server logs
    Week 3–4First Perplexity citations appear in direct query tests
    Week 6–8ChatGPT citations appear for low-competition queries
    Week 10–12Consistent citations across multiple queries and platforms

    If no citations after 12 weeks, check:

    • GPTBot is blocked in robots.txt
    • No schema markup on published pages
    • AEO Score below 75 on published articles
    • Target queries are too competitive for current domain authority
    • Pages have not been submitted to GSC for indexing