Tutorial: Your First GEO Audit

This step-by-step audit will evaluate how well your ecommerce store is optimized for AI search engines. You will check eight critical areas and score your site on a 0-100 scale. Set aside 60-90 minutes, open your store in a browser, and work through each step.

Here is why this matters right now: the average website scores just 31% on GEO readiness, according to the 2026 Conductor AEO/GEO Benchmarks Report. Traditional SEO scores average 72%, meaning most sites have done the classic work but are completely unprepared for AI-driven search. Since AI-generated traffic to U.S. retail sites increased 4,700% year-over-year as of July 2025, the gap between GEO-ready stores and everyone else is widening fast.

Brands appearing in AI-generated answers experience a 38% click lift and a 39% increase in paid ad click performance. Meanwhile, 60% of searches now result in zero-click outcomes. The stores that show up in AI responses capture the demand. The ones that do not lose it entirely.

Step 1: Check Structured Data Coverage (15 points)

Structured data tells AI engines what your pages contain in machine-readable format. Without it, AI engines have to guess — and they often guess wrong or skip your content entirely.

The data here is unambiguous: 71% of pages cited by ChatGPT and 65% of pages cited by Google AI Mode include structured data. Despite this, only about 12.4% of all registered domains have implemented schema.org markup, and the Product schema type appears on just 0.77% of pages globally (HTTP Archive Web Almanac, 2024). Pages with complete product schema are 2.5x more likely to be cited in Google AI Overviews. Schema markup adoption increased 35% between 2023 and 2026, but the vast majority of ecommerce stores still lack complete coverage.

How to test:

Open the Google Rich Results Test at search.google.com/test/rich-results
Enter your homepage URL and run the test
Repeat for one product page, one collection page, and one blog post
Alternatively, use Otterly.ai's GEO Crawlability Checker, which evaluates structured data as part of its 25+ factor audit

What to look for:

Product pages should have Product schema with price, availability, and rating
Blog posts should have Article schema with author and dateModified
FAQ sections should have FAQPage schema (pages with FAQ sections correlate with 4.9 citations versus 4.4 without them)
All pages should have BreadcrumbList schema
Your homepage should have Organization schema

Scoring:

15 points — All five schema types detected across relevant pages
10 points — Product and Article schema present, others missing
5 points — Only basic Product schema present
0 points — No structured data or validation errors

Since only 12.4% of domains have any schema markup at all, scoring even 10 points here puts you ahead of most competitors.

Step 2: Test LLMs.txt (10 points)

LLMs.txt is an emerging standard that helps AI engines understand your site's purpose, structure, and key content. Adoption is growing rapidly from a small base — over 844,000 websites have implemented llms.txt according to BuiltWith tracking as of October 2025. However, an independent analysis of nearly 300,000 domains found only 10.13% had an llms.txt file, and a scan of the top 1,000 websites found zero implementations. The standard has achieved rapid penetration in developer tools, AI companies, and SaaS platforms while remaining largely absent from mainstream ecommerce.

It is worth noting that Google's John Mueller stated in June 2025 that no major AI system currently fetches llms.txt files for grounding — consumer LLMs fetch your actual pages instead. That said, the standard is evolving, and early adoption positions your store for when AI crawlers do begin using it. The cost of implementation is near zero (it is a single text file), so the risk-reward ratio favors adding it.

How to test:

Navigate to yourstore.com/.well-known/llms.txt in your browser
If not found, try yourstore.com/llms.txt
Check if the file exists and contains meaningful content

What to look for:

A good LLMs.txt file should include:

A clear description of your business and what you sell
Links to your most important content pages (guides, FAQs, category pages)
Any brand guidelines for how AI should describe your products
Your product categories and key offerings

Scoring:

10 points — Comprehensive LLMs.txt with business description, key pages, and guidelines
5 points — Basic LLMs.txt exists but is incomplete
0 points — No LLMs.txt file found

Step 3: Evaluate Content Depth (15 points)

AI engines prefer comprehensive content that fully answers user queries. Thin content rarely gets cited. The data backs this up strongly.

Pages above 20,000 characters average 10.18 citations each versus just 2.39 for pages under 500 characters. Content featuring 1,500+ words performs significantly better in AI citation rates, with optimal section length between 100 and 150 words per heading. Material that includes citations, statistics, and quotations achieves 30-40% higher visibility in AI responses. Readable text at a Flesch-Kincaid Grade Level of 6-8 earns 4.6 citations on average versus 4.0 for content written at Grade 11 or above.

The top 4.8% of URLs cited repeatedly by ChatGPT (those cited 10 or more times) are in-depth pages that answer "what is it," "who uses it," "how to choose," and "pricing" all within a single URL. Position 1 pages in traditional search average approximately 1,750 words, while positions 2 through 5 average around 1,500.

Where AI citations pull from within a page also matters: 44.2% of all LLM citations come from the first 30% of text (the introduction), 31.1% from the middle section (30-70%), and 24.7% from the conclusion. This means your opening paragraphs must be substantive and information-dense.

How to test:

Open your top 5 content pages (guides, blog posts, FAQ pages)
For each page, evaluate:
- Word count — Use a browser extension or paste content into a word counter
- Heading structure — Right-click, view source, and search for <h2 and <h3 tags
- FAQ presence — Does the page include a dedicated FAQ section?
- Lists and tables — Are comparisons presented in structured formats?
- Readability — Paste a sample into a Flesch-Kincaid calculator; aim for Grade 6-8

Benchmarks for ecommerce content:

Buying guides: 1,500-3,000 words with 6+ H2 sections
Product descriptions: 300-800 words with 3+ H2 sections
FAQ pages: 15+ questions with detailed answers (50-150 words each)
Blog posts: 800-2,000 words with 4+ H2 sections
Comparison pages: 2,000+ words covering "what," "who," "how to choose," and "pricing" in one URL

Content with structured formatting — hierarchical headings, bullet points, lists, and tables — is 28-40% more likely to be cited by AI engines. Pages with well-organized headings are 2.8x more likely to earn citations in AI search results.

Scoring:

15 points — Most content meets benchmarks, strong heading structure, FAQs present, readability at Grade 6-8
10 points — Some content meets benchmarks, inconsistent structure
5 points — Content exists but is generally thin (under 500 words, few headings)
0 points — Minimal content beyond basic product listings

Step 4: Check Page Speed (10 points)

Slow pages may not fully load for AI crawlers, and speed is a quality signal. As of 2025, only 48% of mobile websites pass all three Core Web Vitals — up from 44% in 2024 and 36% in 2023. Ecommerce sites tend to perform worse than the average due to heavy imagery, third-party scripts, and complex product pages.

The individual metric pass rates tell the story: only 62% of mobile pages achieve a good LCP (the hardest Core Web Vital to pass), while 77% of mobile pages achieve a good INP. For ecommerce specifically, the business impact is severe — revenues nearly double when average page load times drop from 2 seconds to 1 second, and a one-second delay in load time reduces conversions by 7%. For a store generating $100,000 per month, that equals $7,000 in lost revenue every month from speed alone.

How to test:

Open PageSpeed Insights at pagespeed.web.dev
Test your homepage, one product page, and one content page
Record the performance score and Core Web Vitals for each

Key metrics and their thresholds:

| Metric | Good | Needs Improvement | Poor | |---|---|---|---| | LCP (Largest Contentful Paint) | Under 2.5s | 2.5s-4.0s | Over 4.0s | | INP (Interaction to Next Paint) | Under 200ms | 200ms-500ms | Over 500ms | | CLS (Cumulative Layout Shift) | Under 0.1 | 0.1-0.25 | Over 0.25 |

Industry averages for context: Only 48% of mobile pages pass all three metrics simultaneously. If your store passes all three on mobile, you are already in the top half of the web.

Scoring:

10 points — All pages score 80+ with green Core Web Vitals
7 points — Scores between 60-80, one or two orange metrics
3 points — Scores between 40-60, multiple failing metrics
0 points — Scores below 40 or pages timing out

Step 5: Verify Hreflang and Multi-Language (10 points)

If you sell internationally, proper language tagging helps AI engines serve the right version of your content to users in different languages. The cross-border ecommerce consumer market is worth an estimated $1.21 trillion in 2025, with an annual growth rate of 8.71%. Around 28% of sales for international sellers come from cross-border transactions.

The user demand for localization is clear: 75% of international shoppers want to buy products online in their native language. A Seer Interactive case study found that proper hreflang implementation led to approximately 150% higher indexation rates — helping search engines surface the right pages to the right users. For AI engines that serve users globally, correct hreflang signals which language version to reference in a response.

How to test:

View source on your homepage and search for hreflang
Check that each language version of your site has correct hreflang tags
Verify that hreflang tags point to valid, accessible URLs
Confirm that the x-default value is set

What to look for:

<link rel="alternate" hreflang="en" href="https://yourstore.com/" />
<link rel="alternate" hreflang="de" href="https://yourstore.com/de/" />
<link rel="alternate" hreflang="x-default" href="https://yourstore.com/" />

Scoring:

10 points — Correct hreflang on all pages, all languages covered, x-default set
5 points — Hreflang present but incomplete or with errors
2 points — Only on homepage, not on product or content pages
0 points — No hreflang tags (if you run a single-language store, give yourself 10 points and move on)

Step 6: Test Internal Linking Depth (10 points)

AI crawlers discover content through internal links. If important pages are buried deep in your site structure, they may never get crawled or cited. The benchmark across all major SEO sources for 2026 is consistent: key pages should be reachable in 3 clicks or fewer from the homepage.

Sites with strong internal linking structures see measurable gains. Every major ranking boost in 2024-2025 came from reorganizing internal structure rather than acquiring new external links. For AI engines specifically, well-organized headings and clear internal linking make pages 2.8x more likely to earn citations. High-traffic sites (exceeding 1.16 million monthly visitors) average 6.4 citations per query versus 2.4 for low-traffic sites — and internal linking is one of the primary mechanisms that concentrates traffic and authority on key pages.

How to test:

Start from your homepage
Navigate to your most important content page (top guide or FAQ hub)
Count how many clicks it takes to reach it
Check whether your content pages link to related products and vice versa
Verify your navigation menu includes links to content sections (not just product categories)
Use a crawler like Screaming Frog or Sitebulb to map click depth across your entire site

Key indicators:

Click depth — Key content should be reachable in 3 clicks or fewer from the homepage
Cross-linking — Product pages should link to related guides; guides should link to relevant products
Navigation visibility — Content hubs should be accessible from the main navigation
Footer links — Important content categories should appear in the footer
Orphan pages — No important page should exist without at least one internal link pointing to it

Scoring:

10 points — All key content within 3 clicks, strong cross-linking, content in main navigation
7 points — Most content within 3 clicks, some cross-linking present
3 points — Content exists but is isolated from product pages and main navigation
0 points — Content is buried (4+ clicks deep) with no cross-linking

Step 7: Check AI Engine Citations (15 points)

The ultimate test: are AI engines actually citing your store? This step tells you whether everything else is working.

Some context on what to expect: the top 5 domains capture 38% of all AI citations, the top 10 secure 54%, and the top 20 command 66%. Citation distribution is heavily concentrated at the top. On average, AI referral traffic accounts for just 1.08% of all website traffic, though 87.4% of that AI referral traffic comes from ChatGPT. The gap between cited and uncited brands is significant — sites scoring above 70% in GEO readiness receive 4x more AI citations than those below 40%.

ChatGPT now has over 800 million weekly active users (as of October 2025), and Perplexity processes 780 million monthly queries. These are not niche channels anymore.

How to test:

Open ChatGPT (free tier works) and ask questions related to your products:
- "What's the best [your product category] for [common use case]?"
- "Where can I buy [your brand name] [product]?"
- "[Your brand name] reviews"
Repeat the same queries in Perplexity at perplexity.ai
Search for your brand in Google and check if AI Overviews mention you (AI Overviews now appear in 25.11% of Google searches, up from 13.14% in March 2025)
Use a dedicated tracking tool like Otterly.ai (starting at $29/month) or Naridon to monitor citations over time rather than relying on manual spot checks

What to record:

Is your brand mentioned in any AI responses?
Are specific products or pages cited with links?
Is the information about your brand accurate?
How do you compare to competitors in the same responses?

Note: only 38% of URLs cited by Google's AI Overviews come from pages ranking in the top 10 organic results (down from 76% in mid-2025). Traditional SEO rankings no longer guarantee AI citation visibility. You may rank well organically and still be absent from AI responses.

Scoring:

15 points — Cited in multiple AI engines with accurate information and links
10 points — Cited in at least one AI engine with mostly accurate information
5 points — Brand is mentioned but without links or with inaccurate details
0 points — No mentions in any AI engine tested

Step 8: Competitor Comparison (15 points)

Understanding where competitors stand helps prioritize your improvements. Since most competitors score below 35% on GEO readiness (Conductor, 2026), even basic optimization puts you dramatically ahead.

How to test:

Identify your top 3 competitors
Run the same AI engine queries from Step 7 but with competitor brand names
Check one competitor's site for structured data using the Rich Results Test
Compare their content depth to yours on similar topics
Check if competitors have an llms.txt file at their domain

What to evaluate:

Are competitors being cited more frequently than you?
Do they have more comprehensive structured data?
Is their content deeper and better structured?
Do they have features you are missing (LLMs.txt, FAQ schema, comparison pages)?

Remember the concentration effect: the top 5 domains capture 38% of AI citations. If a competitor has established themselves in that top tier for your product category, displacing them requires more than incremental improvements — you need to surpass them on content depth, structured data, and authority simultaneously.

Scoring:

15 points — You are on par with or ahead of competitors across all areas
10 points — Competitors have a slight edge in 1-2 areas
5 points — Competitors are significantly ahead in content and citations
0 points — Competitors dominate AI citations while you are absent

Your GEO Audit Scoring Rubric

Add up your scores from all eight steps:

| Score Range | Rating | What It Means | |---|---|---| | 80-100 | Excellent | Your store is well-optimized for AI search. Focus on maintaining freshness and expanding content. You are in the top tier — pages updated within 2 months earn 5.0 citations on average versus 3.9 for older content. | | 60-79 | Good | Strong foundation with clear areas for improvement. Prioritize the lowest-scoring steps. You are ahead of the 31% average. | | 40-59 | Fair | Significant gaps in AI readiness. Start with structured data and content depth — these two areas drive the largest citation gains. | | 20-39 | Needs Work | You are near the industry average of 31%. Major technical and content improvements needed. Begin with Steps 1, 3, and 4. | | 0-19 | Starting Out | AI optimization has not been addressed yet. The good news: since most competitors score below 35%, focused work over 6-12 weeks can move you ahead quickly. |

Maximum possible score: 100 points

Step 1 (Structured Data): 15 points
Step 2 (LLMs.txt): 10 points
Step 3 (Content Depth): 15 points
Step 4 (Page Speed): 10 points
Step 5 (Hreflang): 10 points
Step 6 (Internal Linking): 10 points
Step 7 (AI Citations): 15 points
Step 8 (Competitor Comparison): 15 points

Tools for Ongoing GEO Monitoring

Manual audits are a starting point, but tracking your GEO performance over time requires dedicated tooling. The GEO tools market has attracted over $31 million in venture funding with 20+ dedicated platforms now available:

Otterly.ai — AI search monitoring across ChatGPT, Perplexity, and Google AI Overviews. Includes a GEO Audit with 25+ evaluation factors. Used by 20,000+ marketing professionals. Starts at $29/month.
Naridon — AI visibility tracking with automated scanning, citation intelligence, and fix suggestions. Includes AEO scoring and competitor monitoring with Shopify integration.
Rankability — GEO research and content optimization platform with llms.txt tracking and adoption data.
SEOscore.tools — Comprehensive audit checklists covering SEO, AEO, and GEO in a single workflow.

Timeline: When to Expect Results

Initial citation visibility typically appears within 6-12 weeks after structured implementation and external brand amplification. Here is the typical progression:

Weeks 1-2: Implement structured data, create llms.txt, optimize content structure. These are the fastest wins.
Weeks 3-6: AI crawlers re-index your updated pages. Adding statistics to your content can improve visibility by up to 40%. Basic, unoptimized content scores around 19.3 in visibility metrics — adding authoritative citations, statistics, and improving fluency pushes scores above 40, a gain exceeding 100%.
Weeks 6-12: Initial citations begin appearing. Monitor with Otterly.ai or Naridon rather than manual spot-checking.
Months 3-6: Compound effects as AI engines build confidence in your content. Sites that audit regularly average 40% more organic traffic, rank for 3x more keywords, and are 5x more likely to appear in AI-generated results.

Re-run this audit every 90 days to track your progress. AI search is evolving rapidly — AI-referred sessions jumped 527% year-over-year in the first five months of 2025, and 98% of CMOs are investing in AEO in 2026. The stores that audit and iterate regularly will maintain their advantage over those that set and forget.