Skip to main content
CitationDesk

For SEO teams

How SEO teams add AI citations to their reporting.

An SEO team adds AI citations to its reporting in four steps: score key pages for citation readiness, fix crawler access, schema and extractability, report clicked ChatGPT citations from GA4 source chatgpt.com, and track citation share per engine with a fixed prompt set.

Rankings and citations are different measurements. A page can hold a top position on Google and still not be the source an AI answer quotes, and the reasons are usually technical ones your team already knows how to fix.

The workflow

How does an SEO team add AI citations to its reporting?

  1. Step 1

    Audit your key pages with the free score

    Paste one URL into the Citation Readiness Score for a 0-100 score and the fix to make first, or give the site audit a sitemap or URL list to score up to 25 pages at once. No signup. Start with pages that already rank for questions your buyers ask, one per template.

  2. Step 2

    Fix robots.txt, schema and extractability

    Allow AI search crawlers such as OAI-SearchBot and PerplexityBot (the AI bot allowlist lists them), add WebPage or Article, Organization and Person schema, and put a direct answer in the first paragraph under question-style H2s. The Fix Kit ranks the issues worst-first. The methodology explains every check.

  3. Step 3

    Report clicked ChatGPT citations from GA4

    When a reader clicks a ChatGPT citation, GA4 records the visit with source chatgpt.com. Break that source down by landing page and it becomes a line in your monthly report. How to check whether ChatGPT cites your site.

  4. Step 4

    Track citation share per engine

    Clicks miss the citations nobody clicked. Run a fixed set of prompts on a schedule, per engine, and count how often your pages are cited. Keep the prompt set and dates in the report so the number can be compared month to month. How to measure AI citation share.

Want the tracking done for you? CitationDesk Pro is a weekly report, €29/mo at the founding price. It finds citations of your pages through Gemini grounding and OpenAI web search (Perplexity is being added, not in reports yet) and lines them up with your Search Console clicks once you connect it.

What we hear from seo teams

What patterns are showing up across your peers?

Rank trackers report positions, not citations

A position report tells you where a URL ranks on Google. It does not tell you whether ChatGPT, Gemini or Perplexity cite that URL when they answer the same question. Some SEO suites now sell AI-visibility add-ons, but the rankings report you already send does not cover it.

A good ranking does not mean the page gets quoted

An AI answer cites a handful of sources. A page can rank well and still be skipped: an AI search crawler is blocked in robots.txt or at the edge, the answer sits deep in the page, or no schema says who published it. Those causes can be checked page by page.

Leadership wants an AI number in the report

Sooner or later someone asks how the site shows up in AI search. Clicked citations are already in GA4, and citation share can be measured with a fixed set of prompts. Both need a stated method before they go in a deck.

How CitationDesk fits

Built for the way seo teams work.

An audit in terms you already use

The score has an AI pillar, a Search pillar and a Convert pillar, and the 0-100 Visibility Score weights them 50/25/25. The checks read like a technical audit: title, canonical, H1 and H2s, WebPage or Article schema, Organization and Person schema, robots.txt and llms.txt.

Up to 25 pages in one pass

The free site audit takes a sitemap or a list of URLs and scores up to 25 pages, so you can see which template has the weakest pillar. No signup, and nothing about your site is stored.

Bot access checked two ways

robots.txt is read separately for AI search crawlers (OAI-SearchBot, PerplexityBot, Claude-SearchBot) and training crawlers (GPTBot, ClaudeBot, CCBot). The audit then fetches your page as PerplexityBot, which catches a block at the edge or WAF that robots.txt does not show.

Worked example

HoldLens.

HoldLens tracks superinvestor positions from quarterly 13F filings. The case study lists the five fixes it made and reports chatgpt.com going from 0 to 10% of its inbound traffic in 14 days. It is one site, so read it as a worked example, not a benchmark.

Read the case study →

FAQ

Questions SEO teams ask

How do SEO teams track AI citations?
With two numbers. Clicked citations come from GA4: sessions with source chatgpt.com, broken down by landing page. Citation share comes from running a fixed set of prompts on a schedule and counting how often each engine cites your pages. State the prompt set, the engine and the dates next to the number.
Does blocking GPTBot stop ChatGPT from citing our pages?
Not on its own. GPTBot is a training crawler. ChatGPT search fetches pages with OAI-SearchBot and ChatGPT-User. The Citation Readiness Score reads robots.txt for the two groups separately and flags a site-wide block of the search crawlers as the one that rules out citation.
What does the Citation Readiness Score check?
Technical SEO basics (title, meta description, canonical, H1 and H2s, WebPage or Article schema, sitemap in robots.txt), entity signals (Organization and Person schema, sameAs), extractability (a short first paragraph, question-style H2s), bot access (robots.txt, llms.txt and a live PerplexityBot fetch) and conversion signals. It measures readiness, not whether you are cited today.
Is it free, and is there an API?
The score and the site audit are free with no signup. The site audit scores up to 25 pages per run. There is no public API: the API reference on this site is an archived design that was never launched.

Get your free Citation Readiness Score.

Paste any URL. 90 seconds. 0-100 score across six dimensions plus the highest-leverage fix you can ship this week.

Or skip pasting a URL — email me a free sample AI-visibility report instead.