Most ecommerce stores are completely invisible to AI systems — and a structured audit is the fastest way to find out if yours is one of them. This checklist gives you 30 items across four categories, each worth points toward a 0-100 AI Visibility Score. Go through it item by item, tally your score, and you'll know exactly where your store stands and what to fix first.
This isn't theory. It's the same framework we use internally to evaluate stores for Generative Engine Optimization (GEO) readiness. If you want the automated version, run your domain through our AI Authority Checker — it takes 30 seconds. But the manual audit below will teach you why each factor matters, which is what you need to actually improve your score.
How the Scoring Works
The audit covers four categories. Each item is scored as Pass (full points), Partial (half points), or Fail (zero). Add them up for your total.
| Category | Max Points | What It Covers | Why It Matters for AI |
|---|---|---|---|
| Technical | 25 | robots.txt, structured data, site speed, crawlability | AI bots must be able to access and parse your content |
| Content | 25 | Blog authority, FAQ pages, comparison content, product depth | AI needs authoritative, structured content to cite you |
| Off-Site | 30 | Reddit, YouTube, review sites, forums, Wikipedia | Third-party mentions drive the majority of AI citations |
| Brand | 20 | Branded search volume, social proof, consistency | AI recommends brands it "trusts" from diverse signals |
| Total | 100 |
Off-Site carries the most weight on purpose. AI systems don't just read your website. They synthesize information from across the internet — YouTube, Reddit, forums, review sites. A store with zero external brand mentions but perfect on-site SEO will still be invisible to AI. The difference between GEO and SEO is exactly this: GEO is about what the rest of the internet says about you, not what you say about yourself.
Category 1: Technical (25 Points)
Technical issues are the easiest to fix and the most common reason stores fail before they even start. If AI bots can't crawl your site or can't parse your data, nothing else matters.
1. robots.txt allows AI crawlers (5 pts)
Check your robots.txt file for directives that block AI user agents: GPTBot, Google-Extended, ClaudeBot, PerplexityBot, Bytespider. If any of these are disallowed, AI systems literally cannot index your content. Many Shopify themes and apps add these blocks without store owners knowing. Pass = all AI bots allowed. Partial = some blocked. Fail = all blocked or no robots.txt.
2. Product schema markup (5 pts)
Every product page should have Product schema with name, description, price, availability, brand, and aggregate rating. This structured data makes your products machine-readable. AI systems extract schema data far more reliably than they parse unstructured HTML. Pass = all product pages have complete Product schema. Partial = some pages or incomplete schema. Fail = no Product schema.
3. FAQ schema on key pages (5 pts)
FAQ schema tells AI exactly what questions your page answers. This directly maps to the conversational queries people type into ChatGPT. Every product page, category page, and your homepage should have FAQ schema. Pass = FAQ schema on 10+ pages. Partial = FAQ schema on 1-9 pages. Fail = no FAQ schema anywhere.
4. Site speed under 3 seconds (5 pts)
AI crawlers have timeouts. If your site takes too long to load, bots abandon the crawl before indexing your content. Test with Google PageSpeed Insights — you want a Largest Contentful Paint under 2.5 seconds and a Time to Interactive under 3.8 seconds. Pass = LCP under 2.5s. Partial = LCP 2.5-4s. Fail = LCP over 4s.
5. Organization schema on homepage (5 pts)
Organization schema (with name, URL, logo, social profiles, and contact info) tells AI systems who you are at the entity level. This is how AI builds a "knowledge card" about your brand. Without it, AI has to infer your brand identity from scattered signals. Pass = complete Organization schema with socials. Partial = basic Organization schema. Fail = none.
Category 2: Content (25 Points)
Content is what gives AI something to cite. Thin product pages with three bullet points give AI nothing to work with. Deep, authoritative content that answers specific questions is what gets recommended.
6. Blog with 20+ posts in your product category (5 pts)
AI systems treat topical depth as an authority signal. A store that has published 20+ substantial blog posts about its product category signals domain expertise. Generic marketing posts don't count — these need to be informational, answering real questions buyers have. Pass = 20+ relevant posts. Partial = 5-19 posts. Fail = fewer than 5 or no blog.
7. Comparison/vs pages (5 pts)
"Product A vs. Product B" is one of the most common queries people ask AI. If you have comparison pages that pit your product against competitors (honestly, with real data), you're directly answering the questions AI gets asked. Pass = 5+ comparison pages. Partial = 1-4 pages. Fail = none.
8. FAQ pages with category-specific questions (5 pts)
Dedicated FAQ pages (not just FAQ schema) give AI long-form answers to reference. These should cover buying decisions, product differences, use cases, and maintenance — the questions a knowledgeable friend would answer. Pass = dedicated FAQ page(s) with 20+ questions. Partial = 5-19 questions. Fail = no FAQ content.
9. Product descriptions over 300 words with specs (5 pts)
One-line product descriptions are invisible to AI. Each product page needs at least 300 words of descriptive content including specific measurements, materials, use cases, and differentiators. AI needs data points to make comparisons. Pass = top 20 products all have 300+ words. Partial = some do. Fail = most under 100 words.
10. Buying guides or how-to content (5 pts)
"How to choose the right [product]" is a high-intent query AI fields constantly. If your store publishes buying guides with clear criteria, score breakdowns, and honest recommendations, AI has structured content to draw from. Pass = 3+ buying guides. Partial = 1-2 guides. Fail = none.
Want the automated version of this audit?
Our free AI Authority Checker scans your store across the signals AI systems actually use and returns a score with a detailed breakdown. Takes 30 seconds. No signup.
Run Your Free AI Audit →Category 3: Off-Site Signals (30 Points)
This is where most stores fail — and where the biggest gains are. AI systems weigh third-party mentions heavily because they're harder to fake than on-site optimization. BrightEdge research shows YouTube alone accounts for 39.2% of AI citation sources, and Reddit has signed $130M+ in AI training data licensing deals. If your brand isn't mentioned on these platforms, AI has very little to work with.
11. YouTube presence — brand mentioned in videos (6 pts)
Search YouTube for your brand name. Are there review videos, unboxing videos, or comparison videos that mention you? YouTube transcripts are heavily indexed by AI models. Pass = 10+ videos mentioning your brand. Partial = 1-9 videos. Fail = zero YouTube presence.
12. Reddit mentions in relevant subreddits (6 pts)
Search Reddit for your brand name. Are people recommending you in relevant threads? Reddit data is literally in AI training sets ($60M Google deal, $70M+ OpenAI deal). One upvoted Reddit recommendation can carry more weight than dozens of blog backlinks in the AI context. Pass = 10+ genuine mentions. Partial = 1-9 mentions. Fail = zero or only self-promotional spam.
13. Review site presence (6 pts)
Does your brand appear on review aggregators — Trustpilot, G2, Capterra (for SaaS), niche review sites for your category? AI pulls from review sites to gauge sentiment and authority. Pass = listed on 3+ review platforms with 50+ total reviews. Partial = 1-2 platforms or fewer reviews. Fail = not on any review site.
14. Third-party editorial mentions (6 pts)
Has your brand been mentioned in editorial content — blog roundups, news articles, industry publications? Search Google for your brand name minus your own domain. These independent citations are trust signals AI systems rely on. Pass = 10+ editorial mentions. Partial = 1-9 mentions. Fail = none.
15. Forum and community presence (6 pts)
Beyond Reddit, are you mentioned on Quora, niche forums, Facebook groups, or other community platforms? The broader your brand's footprint across community discussions, the more confident AI systems are in recommending you. Pass = active presence on 3+ communities. Partial = 1-2 communities. Fail = no community presence.
Category 4: Brand Signals (20 Points)
Brand signals tell AI systems whether you're a real, trusted business or a fly-by-night operation. AI models synthesize these signals to decide confidence level when making recommendations.
16. Branded search volume exists (5 pts)
Do people Google your brand name? Use Google Trends or any keyword tool to check. If your brand name generates zero search volume, AI systems interpret this as "nobody knows this brand." Pass = measurable branded search volume. Partial = very low volume (under 100/mo). Fail = zero branded searches.
17. Consistent brand name across all platforms (5 pts)
Is your brand name spelled the same way on your website, social media, review sites, and marketplace listings? AI systems struggle to connect "BrandName," "Brand Name," and "brand_name" as the same entity. Inconsistency fragments your AI identity. Pass = identical brand name everywhere. Partial = minor variations. Fail = significantly different names across platforms.
18. Social proof signals (5 pts)
Social media followers, engagement, and verified accounts give AI confidence that your brand is legitimate. This isn't about vanity metrics — it's about the signal that real people interact with your brand. Pass = active social accounts with real engagement on 3+ platforms. Partial = 1-2 platforms or low engagement. Fail = no social presence or inactive accounts.
19. Customer review volume and recency (5 pts)
AI systems factor in review freshness. A brand with 500 reviews from 2022 carries less weight than a brand with 100 reviews from the last 6 months. Recent reviews signal an active, current business. Pass = 50+ reviews in the last 6 months. Partial = some recent reviews. Fail = no reviews or all reviews are over a year old.
Scoring Rubric: What Your Total Means
Add up your points from all four categories. Here's what your total score tells you:
| Score | Rating | What It Means | Priority Action |
|---|---|---|---|
| 0 – 20 | Invisible | AI systems have almost no data on your brand. You will not be recommended. | Fix technical blockers (robots.txt, schema), then focus entirely on building off-site mentions. |
| 21 – 40 | Weak | AI might know you exist but lacks confidence to recommend you. | Build YouTube and Reddit presence. Publish 5-10 deep content pieces. Get listed on review sites. |
| 41 – 60 | Developing | AI recommends you for some queries but not consistently. Gaps remain. | Fill content gaps. Increase editorial coverage. Expand community presence to 3+ platforms. |
| 61 – 80 | Strong | AI recommends you regularly. You're a recognized player in your category. | Optimize for adjacent queries. Defend against emerging competitors. Keep content fresh. |
| 81 – 100 | Dominant | AI consistently names you as a top recommendation. You own your category in AI. | Expand into adjacent categories. Monitor for new AI platforms. Maintain review freshness. |
If you scored under 40, that's normal. Most ecommerce stores we've evaluated land in the 10-30 range. The reason is simple: most store owners have never thought about AI visibility as a channel. They've been focused on Google rankings and paid ads. But as we covered in our GEO vs. SEO breakdown, 88% of AI-cited URLs don't rank in Google's top 10 (BrightEdge). These are separate games.
The Quick-Start Fix Priority
You don't need to fix everything at once. Here's the order that moves your score fastest, based on the point values and typical effort required:
| Priority | Action | Points Available | Effort | Timeline |
|---|---|---|---|---|
| 1 | Unblock AI crawlers in robots.txt | 5 | 5 minutes | Today |
| 2 | Add Product + FAQ schema to top 10 pages | 10 | 2-3 hours | This week |
| 3 | Create 3 comparison/vs pages | 5 | 6-8 hours | This week |
| 4 | Write 5 genuine Reddit comments in niche subs | 6 | 1 hour | This week |
| 5 | Get listed on 2 review platforms | 6 | 1-2 hours | This week |
| 6 | Record or commission 1 YouTube review | 6 | 2-4 hours | This month |
| 7 | Reach out to 10 bloggers for roundup inclusion | 6 | 3-4 hours | This month |
| 8 | Publish 10 deep blog posts in your category | 5 | 20-30 hours | Next 30 days |
Following this priority order, you can gain 20-30 points in the first week. That's enough to move most stores from "Invisible" to "Weak" or "Developing." The technical fixes (priorities 1-2) are quick wins that remove blockers. The off-site work (priorities 4-7) builds the external signals AI systems weigh most heavily.
Common Audit Failures We See
After evaluating stores across multiple categories, patterns emerge. Here are the most common failures:
Blocking AI crawlers without knowing it. Many Shopify apps and themes add Disallow directives for GPTBot and other AI user agents to robots.txt. Store owners never check because traditional SEO tools don't flag it. One line in robots.txt can make your entire store invisible to ChatGPT.
Strong SEO but zero off-site presence. Stores that rank well on Google often assume they're visible to AI too. They're not. A store ranking #1 for its main keyword with zero Reddit mentions, no YouTube reviews, and no third-party editorial coverage will score under 30 on this audit.
Thin product pages. Most Shopify stores have product descriptions under 50 words. That's not enough for AI to understand what makes your product different from competitors. AI needs specific claims, measurements, comparisons, and use cases to make informed recommendations.
Inconsistent brand naming. "MyBrand" on the website, "My Brand Store" on Instagram, "mybrand_official" on TikTok. AI systems can't reliably connect these as the same entity. Every variation fragments your brand's AI identity and reduces recommendation confidence.
No comparison content. "What's better, X or Y?" is one of the most common questions people ask AI. If you don't have comparison pages, AI has to rely on third-party sources to answer — and those sources might not recommend you.
How This Audit Connects to Your AI Visibility Score
The score from this manual audit should roughly correlate with the automated score from our AI Authority Checker. The manual version is more granular — it tells you exactly which items need attention. The automated version is faster and tracks changes over time.
Use both. Run the manual audit once to build your improvement roadmap. Then use the automated checker monthly to track progress and catch regressions.
For a deeper understanding of what drives AI recommendations and how to systematically improve your standing, read our guide on how to get your Shopify store recommended by AI.
FAQ
What is a GEO audit?
A GEO audit is a systematic review of your ecommerce store's readiness to be recommended by AI systems like ChatGPT, Perplexity, Gemini, and Claude. It evaluates four areas: technical setup (structured data, robots.txt, site speed), content authority (blog depth, FAQ pages, comparison content), off-site signals (Reddit mentions, YouTube presence, review sites), and brand strength (branded search volume, social proof). The result is a score from 0-100 that tells you how AI-visible your store is.
How do I score my store's AI visibility?
Use the 30-point checklist above. Score each item as Pass (full points), Partial (half points), or Fail (zero). Add up your totals across all four categories: Technical (25 pts), Content (25 pts), Off-Site (30 pts), and Brand (20 pts). For an automated score, use our AI Authority Checker.
What is the most important factor in a GEO audit?
Off-site signals carry the most weight (30 out of 100 points). AI systems rely heavily on third-party mentions to decide which brands to recommend. YouTube accounts for 39.2% of AI citation sources (BrightEdge), and Reddit data has been licensed in $130M+ worth of AI training deals. A store with strong off-site mentions but weak on-site technical setup will still outperform a perfectly optimized site with zero external brand presence.
Should I block AI crawlers in robots.txt?
No. Blocking AI crawlers (GPTBot, Google-Extended, ClaudeBot, PerplexityBot) in your robots.txt prevents those systems from indexing your content. If you want AI to recommend your store, allow these bots. Check your robots.txt right now — many Shopify apps and themes add these blocks without the store owner knowing.
How often should I run a GEO audit?
Run a full manual audit quarterly. Between audits, check your AI Authority Checker score monthly to track progress. AI models retrain frequently, and your competitive landscape shifts as more stores discover GEO.
Can a small store score well on a GEO audit?
Yes. Unlike SEO, where domain authority and backlink history create massive barriers, GEO has no legacy advantage. AI models retrain regularly, so a small store that builds strong brand mentions on YouTube, Reddit, and review sites can score higher than an established brand that hasn't optimized for AI. Every item on this checklist is achievable regardless of store size or budget.