GEO AI Search Audit: 0-100 Score & Instant Code Fixes

I'll audit stripe.com/docs now: fetching the page, its robots.txt, and llms.txt, then scoring it.

The docs redirect to docs.stripe.com, which has its own robots.txt and llms.txt, so I'm checking those next. I'm also trying to get the raw HTML to look for JSON-LD.

I have the data I need. The raw-HTML check was blocked by the sandbox network allowlist, so I'll flag JSON-LD as unverified in the report. Updating the task list before writing it up.

🌍 GEO Audit Report: Stripe Documentation

URL Audited: https://stripe.com/docs (redirects to https://docs.stripe.com) | Date: 2026-10-17

Limitation: I could not inspect the raw HTML. My shell is restricted to package registries, and WebFetch returns pages converted to markdown. The JSON-LD and SSR scores below are therefore estimates, not confirmed findings. To verify, paste the page source (View Source) and I'll rescore Pillar 3.

📊 Overall GEO Score: 66 / 100 — Needs Work

Pillar 1 breakdown: robots.txt 10/10, text readability 5/5, llms.txt 5/5.

🔍 What I Found

docs.stripe.com/robots.txt: Allows everything (Allow: /) for all user agents. It also carries a Content-Signal: ai-train=yes, search=yes, ai-input=yes line. Only /handoff paths are disallowed.

stripe.com/robots.txt (root): The wildcard rule allows /docs. A special rule disallows /docs for ia_archiver only, so the Internet Archive crawler is excluded from the docs path. This is not an AI-bot issue.

stripe.com/llms.txt: Exists, with a clear title, a summary blockquote, and sections for products, solutions and resources. It links to docs.stripe.com/llms.txt.

docs.stripe.com/llms.txt: Exists and is exceptionally thorough. It includes a section of best-practice instructions for LLM agents (preferred APIs, deprecated endpoints to avoid). Pages are available in .md form.

Content visibility: The page comes through as clean text, so core content is not trapped behind client-side JavaScript.

🚨 Top Red Flags Blocking AI Visibility

No definition of "Stripe" on the docs landing page: The page opens with a generic "explore our guides" sentence. An AI engine extracting a passage can't tell what Stripe is, what category it belongs to, or who it serves, so the hub is rarely cited for definitional queries.
2.

No self-contained, fact-dense answer blocks: The hub is almost entirely link lists. There are no question-style headings followed by a 130–165 word summary that can be lifted without context.
3.

Structured data unconfirmed and likely thin: I couldn't verify any JSON-LD (Organization, WebSite or FAQPage) on the page, and there are no visible sameAs links or update dates to anchor the entity.

🛠️ Ready-to-Use Fixes (Copy & Paste)

Fix 1: Maintenance Check: robots.txt

Pillar 1 scored 20/20, so this is a maintenance check rather than a fix. The docs file already allows all bots. If you want per-bot clarity, this keeps your existing Disallow rules and adds explicit named allowances:

Sitemap: https://docs.stripe.com/sitemap.xml

User-agent: GPTBot
Allow: /

User-agent: OAI-SearchBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: Google-Extended
Allow: /

User-agent: Applebot-Extended
Allow: /

User-agent: *
Content-Signal: ai-train=yes, search=yes, ai-input=yes
Allow: /
Disallow: /handoff-healthcheck
Disallow: /handoff

One thing to check: ia_archiver is disallowed from /docs on the root domain. Confirm that is intentional.

Fix 2: Improved /llms.txt (short, definitional header)

Your current file is strong but very long. A short curated top section helps agents orient before the full index:

# Stripe Documentation

> Stripe is a financial infrastructure platform that lets businesses accept online and in-person payments, manage subscriptions and invoices, automate tax, and embed financial services. The documentation covers APIs, SDKs, integration guides and testing for developers.

## Start Here

- [Accept a payment](https://docs.stripe.com/payments/accept-a-payment.md): Integrate online payments using Checkout or Elements.
- [Quickstarts](https://docs.stripe.com/quickstarts.md): Working integration examples by use case.
- [API Reference](https://docs.stripe.com/api.md): Complete reference for all Stripe API endpoints.

## Products

- [Payments](https://docs.stripe.com/payments.md): Accept cards, wallets, bank debits and local payment methods.
- [Billing](https://docs.stripe.com/billing.md): Subscriptions, invoicing and usage-based pricing.
- [Connect](https://docs.stripe.com/connect.md): Payments for platforms and marketplaces.
- [Tax](https://docs.stripe.com/tax.md): Automated sales tax, VAT and GST.

## Optional

- [Changelog](https://docs.stripe.com/changelog.md): API changes and upgrades.
- [Support](https://support.stripe.com): Help and account support.

Fix 3: Citability Rewrite (Before vs. After)

• ❌ Current Weak Passage (docs landing intro):

  "Explore our guides and examples to integrate Stripe."

• ✅ GEO-Optimized Answer Block (about 150 words, paste at top of the page):

What is Stripe and how do I integrate it? Stripe is a financial infrastructure platform for businesses of every size, from startups to public companies, that need to accept payments, manage recurring revenue, and embed financial services. Instead of stitching together separate merchant accounts, gateways and processors, developers integrate once through Stripe's APIs and prebuilt components. The Stripe CLI installs with npm install -g @stripe/cli, and stripe agent setup configures agent-assisted integrations. Core documentation areas include: (1) Payments, covering Checkout, Elements and Payment Links; (2) Billing, covering subscriptions, invoices and usage-based pricing; (3) Connect, covering marketplaces and platform payouts; and (4) Tax, covering automated sales tax, VAT and GST. Stripe supports [Insert number] payment methods across [Insert number] countries and processes [Insert annual payments volume, source from Stripe's annual letter]. Most developers complete a first test payment in [Insert typical timeframe] using the quickstarts.

The CLI command comes from the page itself. The bracketed items are placeholders because I didn't find those figures on the audited page. Fill them from Stripe's official sources rather than estimates.

Fix 4: JSON-LD Structured Data

Add this to the <head> of the docs landing page. If a script already exists, merge rather than duplicate.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "name": "Stripe",
      "url": "https://stripe.com",
      "description": "Stripe is a financial infrastructure platform that lets businesses accept online and in-person payments, manage subscriptions, automate tax, and embed financial services.",
      "sameAs": [
        "https://www.linkedin.com/company/stripe",
        "https://x.com/stripe",
        "https://github.com/stripe"
      ]
    },
    {
      "@type": "WebSite",
      "name": "Stripe Documentation",
      "url": "https://docs.stripe.com",
      "publisher": { "@type": "Organization", "name": "Stripe" }
    },
    {
      "@type": "FAQPage",
      "mainEntity": [
        {
          "@type": "Question",
          "name": "What is Stripe?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Stripe is a financial infrastructure platform for businesses. It provides APIs and tools to accept payments, manage billing, and embed financial services."
          }
        },
        {
          "@type": "Question",
          "name": "How do I start integrating Stripe?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Install the Stripe CLI, then follow a quickstart for Checkout or Elements. Test your integration in a sandbox before going live."
          }
        }
      ]
    }
  ]
}
</script>

Check the sameAs URLs against Stripe's official profiles before publishing.

---

⚡ 3-Step Action Plan (Priority Order)

Today (5 mins): No robots.txt change is needed. Prepend the short definitional header from Fix 2 to /llms.txt so agents get orientation before the full index.
2.

This week (30 mins): Add the JSON-LD from Fix 4 to the docs landing page <head>, and replace the one-line intro with the Fix 3 answer block, with the placeholders filled in.
3.

Ongoing: Add a short question-style FAQ ("How do I accept a payment?", "How do I test my integration?") to the top five docs entry pages.

I scored this site strictly, so the 66 reflects missing citable passages and unverified schema, not any crawler problem. Stripe is unusually well set up for AI crawlers.

Sources

• docs.stripe.com
• docs.stripe.com/robots.txt
• stripe.com/robots.txt
• stripe.com/llms.txt
• docs.stripe.com/llms.txt