GEO Content Auditor — AI Answerability & Citation Readiness

Audit public webpages for AI answerability and citation readiness. This GEO content auditor scores each page from 0–100 and returns a prioritized fix list for clearer structure, extractable facts, entity signals, machine readability, and authority—factors that can help ChatGPT, Perplexity, Claude, and Gemini understand and cite content.

This is the "fix your own pages" piece of the GEO family: the AI Visibility Rank Tracker tells you if you're being mentioned; the AI Citation Monitor shows which sources AI cites; the GEO Content Auditor tells you why your page isn't being picked and exactly what to change.

---

What it audits

Each page is scored across 5 dimensions (100 pts total):

Dimension	Max	What it checks
Structure	30	Question-style headings, direct answer near top, FAQ sections, lists/tables, H1-H3 hierarchy
Extractable facts	25	Specific stats, dates, named entities, clear definitions, concrete examples
Entity clarity	20	Brand clearly defined, author byline, JSON-LD structured data, product clearly named
Machine readability	15	Semantic markup, content chunkability, llms.txt presence
Authority signals	10	Sources cited, publication date, internal links, breadcrumbs

What the GEO audit returns

Output	How to use it
Overall score and letter grade	Compare pages and prioritize the weakest content
Five dimension scores	Identify whether structure, facts, entities, readability, or authority is the main constraint
Strengths and issues	Preserve what works and create an evidence-backed remediation backlog
Quick wins	Start with high-impact changes that are easier to implement
Missing schema types	Plan structured-data improvements such as Organization or FAQPage where appropriate
Suggested Q&A pairs	Align content with target buyer questions without guessing search intent
Site summary	Find recurring issues and priority pages across the audited URL set

---

Output per page

{
  "url": "https://example.com/blog/best-tools",
  "page_title": "Best Productivity Tools 2026",
  "ai_answerability_score": 61,
  "grade": "C",
  "structure_score": 18,
  "facts_score": 15,
  "entity_score": 12,
  "readability_score": 10,
  "authority_score": 6,
  "strengths": ["Clear H2 headings", "Includes specific statistics"],
  "issues": [
    "No FAQ schema — add FAQPage JSON-LD to become directly quotable",
    "No direct 1-sentence answer in opening paragraph",
    "Brand name not defined clearly — add Organization JSON-LD"
  ],
  "quick_wins": [
    "Add FAQPage JSON-LD with 5 buyer questions",
    "Open with a direct 1-sentence answer",
    "Add publication and last-updated dates"
  ],
  "missing_schema_types": ["FAQPage", "Organization"],
  "suggested_qa_pairs": [
    { "question": "What is the best productivity tool for teams?", "answer_hint": "Compare top 3 with use cases" }
  ],
  "llms_txt_present": false,
  "has_json_ld": false,
  "word_count": 1240,
  "prompt_fit_score": 55,
  "error": null,
  "audited_at": "2026-06-08T12:00:00.000Z"
}

Plus a SUMMARY in the key-value store: site-wide average score, top issues across all pages, quick wins, priority pages (lowest scoring = biggest opportunity), and a missing schema frequency table.

---

Input

Field	Type	Description
startUrls	URL list	Pages to audit.
brandName	string	Your brand/product name (entity clarity check). Default: Apify.
targetPrompts	string[]	Buyer questions this content should answer. Optional — scores prompt-fit when provided.
maxPages	integer	Max pages per run (1–50). Default: 10.

The LLM model is managed internally for predictable cost and quality.

Example input

{
  "startUrls": [
    { "url": "https://yoursite.com/blog/best-tools" },
    { "url": "https://yoursite.com/product/features" }
  ],
  "brandName": "YourBrand",
  "targetPrompts": [
    "What is the best tool for X?",
    "How do I solve Y?"
  ],
  "maxPages": 10
}

---

API key (required)

All LLM analysis runs through OpenRouter — set one environment variable in the Apify Console (Settings → Environment variables, mark Secret):

OPENROUTER_API_KEY = sk-or-...

---

Pricing (Pay Per Event)

Event	When
actor-start	Once per run.
page-audited	Each successfully audited page.

Failed pages are not charged.

maxPages places a hard ceiling on the number of page-audit events. OpenRouter model usage is paid separately through your own OpenRouter account. Review the current event prices in the Apify Store before a production run.

Run a GEO content audit through the API

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('USERNAME/geo-content-auditor').call({
  startUrls: [
    { url: 'https://example.com/product' },
    { url: 'https://example.com/blog/category-guide' },
  ],
  brandName: 'Example Brand',
  targetPrompts: ['What is the best tool for category teams?'],
  maxPages: 10,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Use schedules to repeat audits after content releases, webhooks to notify a content workflow, dataset exports for analysis, or Apify MCP for agent-driven optimization tasks.

---

Use cases

• GEO content strategy — find the exact structural and semantic gaps keeping your pages out of AI answers.
• Content audit before publishing — score new content before it goes live.
• Agency client reporting — deliver an AI-readiness score with actionable fixes.
• llms.txt gap detection — instantly see which of your domains are missing it.
• Pairs with AI Visibility Rank Tracker + AI Citation Monitor for the full GEO picture: are you cited? → what do AI engines cite? → why aren't your pages being picked?

When to use—and when not to use—this actor

Use it for page-level GEO audits, AEO content reviews, pre-publication quality checks, agency deliverables, schema-opportunity discovery, and prioritizing an existing content library. Supply targetPrompts when the page is intended to answer specific buyer questions; omit them for a general citation-readiness review.

Do not use it as proof that a model will cite a page, a Google ranking predictor, a full technical SEO crawler, or a substitute for schema validation and human editorial review. The score evaluates observable page signals and LLM analysis; model providers do not publish a universal citation formula.

Best practices for AI answerability audits

• Audit canonical, publicly accessible page URLs rather than preview or login-only links.
• Use specific buyer questions in targetPrompts, not single keywords.
• Compare like-for-like page types instead of treating every score as directly equivalent.
• Re-run the same URLs after implementing changes to measure score movement.
• Validate suggested structured data against the current Schema.org and search-engine rules before publishing it.
• Prioritize recurring issues across several pages over one isolated suggestion.

FAQ

What is AI answerability?

AI answerability describes how easily an answer engine can identify the page's subject, extract a concise response, connect claims to entities and sources, and reuse the information in a generated answer.

Does a high score guarantee ChatGPT or Perplexity citations?

No. Citation selection also depends on query relevance, authority, freshness, model behavior, retrieval systems, and competing sources. The score is a diagnostic benchmark, not a placement guarantee.

What is the difference between GEO and traditional SEO?

SEO commonly focuses on search ranking and clicks. GEO and AEO focus on whether generated answers can understand, mention, and cite the brand or content. Strong technical SEO and useful original content still support both.

Can agencies audit multiple client sites?

Yes. Run separate inputs per client or site, export the dataset, and use the SUMMARY key for portfolio-level reporting. Keep comparable prompts and page types for consistent benchmarking.

Reliability and responsible use

LLM analysis is probabilistic, so review recommendations before changing production content. A fetch error or missing page text is not a valid low GEO score; inspect the error field first. Store OPENROUTER_API_KEY as a secret environment variable. Only audit content you are authorized to access and comply with applicable website, copyright, privacy, and data-protection requirements.