How to detect GPTBot on your WordPress site.

Why GPTBot is invisible to standard analytics

GPTBot is a server-side HTTP client. When it visits your WordPress site, it does this:

GET /your-blog-post/ HTTP/1.1
Host: yoursite.com
User-Agent: Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko;
            compatible; GPTBot/1.2; +https://openai.com/gptbot)
Accept: text/html,application/xhtml+xml

It downloads the HTML response, parses it, extracts text, and leaves. It never executes JavaScript. Which means:

• Your Google Analytics gtag.js never fires → GA shows 0 visits from GPTBot.
• Plausible's tracking script never loads → Plausible shows 0 visits from GPTBot.
• Fathom, Simple Analytics, GoatCounter, Umami, Matomo's JS mode — all blind.

The only way to see GPTBot is to log requests on the server itself: from PHP, Apache logs, or a server-side analytics layer.

Method 1: Apache / Nginx logs (manual, free, ugly)

If you have SSH access, you can grep the access log:

# Apache
tail -f /var/log/apache2/access.log | grep -i gptbot

# Nginx
tail -f /var/log/nginx/access.log | grep -i gptbot

This works but: requires SSH access, only shows raw lines, no aggregation, no dashboard, no historical data unless you also configure a log retention policy. Fine for one-off sanity checks; not a monitoring solution.

Method 2: SysWP Radar plugin (3 minutes, dashboard, alerts)

SysWP Radar's WordPress plugin captures GPTBot (and every other AI crawler) on every request, server-side, regardless of JavaScript. Setup:

1. Create a free account at radar.syswp.com.br/auth/signup. 7-day Pro trial included, no credit card.
2. Add your WordPress site on the Sites page. Copy the 12-character Site ID.
3. Install the plugin: download syswp-radar.zip, go to WP Admin → Plugins → Add New → Upload Plugin → select the zip → Install → Activate.
4. Paste the Site ID: WP Admin → Settings → SysRadar → paste → Save.
5. Done. Open the dashboard at radar.syswp.com.br/dashboard and navigate to the Top AI bots panel.

Within 5-60 minutes of install, depending on your traffic, you'll see your first GPTBot hit alongside ClaudeBot, PerplexityBot, Amazonbot, and other AI crawlers. Each row shows: bot name, hit count, top URIs crawled, country of origin.

What to do with GPTBot data

• Identify high-value content for GEO (Generative Engine Optimization). The pages GPTBot revisits are signals that OpenAI considers them strong answer sources. Double down on similar topics.
• Spot 404 errors GPTBot hits. If GPTBot is trying to fetch /some-old-post/ that's gone, you're losing AI search visibility. Set up a 301 redirect.
• Verify your robots.txt rules. If you've added User-agent: GPTBot + Disallow: /, Radar shows you whether GPTBot is actually respecting it (it usually does).
• Track relative volume vs ClaudeBot, PerplexityBot. Each AI feeds a different user base. Knowing which crawls your content most is a competitive signal.

How to block GPTBot (if you don't want it)

Detection ≠ blocking. If you decide GPTBot isn't welcome:

# /robots.txt
User-agent: GPTBot
Disallow: /

OpenAI publicly commits to honoring this. To verify, watch the SysRadar dashboard after deploying — GPTBot hits should drop to zero within 24-48 hours. If they don't, escalate (rare but possible).

For aggressive blocking (header-level, not opt-in), use a Cloudflare WAF rule or pair Radar with SysWP Shield for one-click blocking.

FAQ