The Anti-GEO Playbook: 15 Mistakes That Make AI Skip Your Brand
Most GEO advice tells you what to do. This playbook tells you what to stop doing. 15 mistakes that make AI platforms skip your brand, ranked by severity, with the evidence behind each one. Every statistic is sourced from a published study.
Most GEO advice tells you what to do. This guide tells you what to stop doing. It is a catalog of 15 specific mistakes that make AI platforms skip your brand, ranked by severity, with the evidence behind each one. Every statistic in this post comes from a published source we verified directly; where a mistake carries no figure, we explain the mechanism instead.
2.9%
of AI citations point to a brand's own site; the rest is what others publish about you (Ranqo, 149,912 citations across 102 brands)
We grouped the 15 mistakes into three severity tiers. Severe mistakes make your brand functionally invisible to AI. Moderate mistakes cause significant citation loss but leave a path to recovery. Subtle mistakes compound over time -- they don't kill visibility on their own, but combined they keep you locked out of high-intent queries. For context on the broader landscape, see Ranqo's complete GEO guide.
All 15 Mistakes at a Glance
The full catalog, with the evidence behind each mistake. The detailed breakdown of every mistake follows below.
All 15 Anti-GEO Mistakes
Every mistake mapped to its severity tier and the evidence behind it
| # | Mistake | Severity | Evidence | Source |
|---|---|---|---|---|
| 1 | JavaScript-only rendering with no SSR | Severe | OpenAI, Anthropic and Perplexity crawlers do not render JavaScript | Vercel |
| 2 | Paywalled or login-gated key content | Severe | A paywall or login hides the text a crawler would quote | Editorial |
| 3 | Blocking AI crawlers in robots.txt | Severe | Blocking Claude's search bots may reduce your visibility in its answers | Anthropic crawler docs |
| 4 | Anonymous corporate content (no bylines) | Severe | Google asks for a byline where readers expect one, in Search and its AI features | Google Search Central |
| 5 | Optimizing only for Google rankings | Severe | Only 12% of AI chatbot citations match Google top 10 | Ahrefs |
| 6 | Thin content on important pages | Moderate | Complete answers give an engine more to quote | Editorial |
| 7 | Missing FAQ / Article / Product schema markup | Moderate | Schema alone barely moves AI citations; it helps when it mirrors visible facts | Ahrefs (1,885 pages) |
| 8 | Outdated content (no update in 12+ months) | Moderate | AI-cited content runs 25.7% fresher than Google's organic results | Ahrefs (17M citations) |
| 9 | No original statistics or proprietary data | Moderate | Statistics addition = about +31% visibility in a simulated AI engine; its prompt allowed hypothetical statistics | Princeton GEO |
| 10 | Optimizing ChatGPT for Google instead of Bing | Moderate | 87% of ChatGPT Search citations match Bing top results; only 56% match Google | Seer Interactive |
| 11 | No "best X" or comparison content | Subtle | Ranked listicles draw 35.7% of content citations; comparison pages 4.5% | Ranqo (149,912 citations) |
| 12 | Prose paragraphs instead of structured tables | Subtle | A table states each fact once, in a form an engine can lift | Editorial |
| 13 | No third-party mentions (review sites, press) | Subtle | Only 2.9% of AI citations point to a brand's own site | Ranqo (102 brands) |
| 14 | One-time optimization with no maintenance | Subtle | 77.5% of brand-question-engine pairs are named every time or never | Ranqo (102 brands) |
| 15 | Single-platform focus (one-size-fits-all) | Subtle | U.S. adults using each: ChatGPT 44%, Gemini 24%, Claude 6%; each uses distinct signals | Pew Research, 2026 |
Tier 1: Severe Mistakes (Functionally Invisible)
These five mistakes don't reduce your AI visibility -- they eliminate it. If any of them apply to your site, fix them before touching anything else.
1. JavaScript-only rendering with no SSR
Per Vercel's direct analysis of AI crawler behavior, none of the major AI crawlers currently render JavaScript. If your content is rendered client-side via React, Vue, or Angular without server-side rendering or static generation, GPTBot, ClaudeBot, and PerplexityBot see an empty shell. Your content does not exist to them.
Fix: enable SSR or static generation for all content pages. Verify by viewing your page with JavaScript disabled in your browser dev tools -- if the content is gone, AI cannot read it.
2. Paywalled or login-gated key content
A crawler that reaches a paywall or a login form reads the gate, not the article, so there is nothing on the page for it to quote or cite. Your most valuable content (pricing, comparisons, key documentation) needs to be publicly indexable.
Fix: identify which pages must be public (pricing, product features, comparisons, top docs, case studies, security/compliance pages) and remove any login or paywall gates.
3. Blocking AI crawlers in robots.txt
OpenAI and Anthropic say their crawlers honor robots.txt. If you have accidentally blocked their search crawlers (OAI-SearchBot, Claude-SearchBot), or worse, blocked them deliberately to "protect" your content, you have opted out of being cited in their answers. Anthropic's crawler documentation spells out the cost: blocking its search bots may reduce your site's visibility in Claude's answers to users.
Fix: audit your robots.txt for explicit blocks of GPTBot, ChatGPT-User, ClaudeBot, Claude-User, PerplexityBot, Google-Extended, or CCBot. Remove all blocks for content you want cited.
4. Anonymous corporate content with no bylines
Google's guidance on helpful content asks whether it is self-evident who wrote a page and whether pages carry a byline where readers would expect one, and Google says the same guidance applies to AI Overviews and AI Mode. A page signed by "[Brand] Team" gives an engine no accountable author to credit.
Fix: add a real named author with title, brief bio, and credentials to every content page. For technical content, the engineering lead authors. For business content, an executive or domain expert authors. "Acme Marketing Team" doesn't count.
5. Optimizing only for Google rankings
Ahrefs analyzed 15,000 prompts and found that only 12% of URLs cited by AI chatbots rank in Google's top 10 for the same query. Your #1 Google ranking does not transfer to AI visibility. For the full breakdown, see Ranqo's research on the Google-AI disconnect.
Fix: add AI visibility KPIs (mention rate, share of voice, position in AI responses) alongside your existing SEO metrics. Track each platform separately.
A single Tier 1 mistake can keep AI from reading your pages at all. Fix these before you optimize anything else.
Tier 2: Moderate Mistakes (Significant Losses)
These five mistakes won't make you invisible, but each one weakens the case for citing your page. Most companies have several of these active simultaneously.
6. Thin content on key pages
A page that answers the whole question gives an engine more to quote. A 600-word article competing with a comprehensive 2,500-word guide on the same question usually loses: the guide covers the follow-up questions the short page leaves open.
Fix: for any high-priority informational page (comparison, category guide, best-of, how-to), cover the question fully, with genuine depth rather than padding. Add specific examples, original data, edge cases, and expert quotes.
7. Missing FAQ, Article, or Product schema markup
The Ahrefs 1,885-page controlled study found schema markup had near-zero independent effect on AI citation rates. Schema is verification, not amplification: the visible Q&A and structured content the markup reflects is what actually gets cited, and the markup makes those facts unambiguous to a machine. It is the mildest item in this tier: hygiene, not a multiplier.
Fix: add JSON-LD schema (FAQ, Article, HowTo, Product) to top 10 content pages. Validate with Google Rich Results Test. Most CMS platforms support schema natively or via plugins.
8. Outdated content (no update in 12+ months)
Ahrefs' study of 17 million citations found AI-cited content is 25.7% fresher than the pages Google ranks organically. ChatGPT is the most extreme, citing pages 393 to 458 days newer than Google's results, and Perplexity appears to list its newest sources first.
Fix: add visible "last updated" dates. Set a quarterly content refresh cadence for top pages. Refresh statistics, examples, and screenshots even when the core thesis hasn't changed.
9. No original statistics or proprietary data
Princeton's peer-reviewed GEO research (KDD 2024) found that adding statistics raised a source's share of a simulated AI engine's answers by about 31%, one of its strongest results. In that study the model was allowed to invent the statistics it added, so use only numbers you can source, as our page refresh checklist explains. Generic content, the same advice available on ten other sites, gives AI no reason to cite yours specifically.
Fix: include at least 3 original statistics or data points per content page. Run your own surveys, analyses, or case studies. Cite specific numbers from your customer base or product analytics.
10. Optimizing ChatGPT for Google instead of Bing
Seer Interactive verified that 87% of ChatGPT Search citations match Bing's top results. Google's match rate is just 56%. ChatGPT, the chatbot 44% of U.S. adults use, is built on Bing, not Google.
Fix: submit your sitemap to Bing Webmaster Tools. Monitor your Bing rankings for key queries. For more platform-specific tactics, see Ranqo's platform-specific playbook.
Tier 3: Subtle Mistakes (Compound Losses)
These five mistakes are easy to overlook because each one's impact is moderate in isolation. But they compound. A site making all five Tier 3 mistakes can lose more total citation surface than one making two Tier 1 mistakes.
11. No "best X" or comparison content
Shortlist questions are answered from ranked lists. In Ranqo's study of 149,912 AI citations, ranked "best X" listicles drew 35.7% of the citations that went to content, against 4.5% for two-way comparison pages. Most brands appear on neither.
Fix: earn a place on the "best [category] for [use case]" lists your buyers' questions pull up, then keep a balanced head-to-head page for each top competitor, with feature-by-feature tables and explicit verdicts ("X is better for use case A; Y is better for use case B").
12. Prose paragraphs instead of structured tables
AI platforms extract labeled rows and columns reliably. A table states each fact once, in a form an engine can lift without guessing. Paragraphs are ambiguous; tables are not.
Fix: for any content that compares options, lists features, or shows specs, convert prose to HTML tables. Aim for 3 columns and 5-6 rows per table.
13. No third-party mentions or review presence
Only 2.9% of the 149,912 AI citations in Ranqo's 102-brand study pointed to a brand's own site; 75.2% went to corporate and competitor pages. Your own website is one input. What others say about you is the dominant signal.
Fix: get listed and verified on G2, Capterra, TrustPilot, or industry-specific review platforms. Earn coverage on review blogs and comparison sites. Build a YouTube presence: YouTube is one of the seven websites every engine cites whatever a company sells.
14. One-time optimization with no ongoing maintenance
Whether an engine names you for a question is mostly settled: across repeated runs in Ranqo's study of 102 brands, 77.5% of brand, question and engine combinations were named every time or never. A brand that is never named stays that way until its sources change, and a rival's new page can move you the other way. Without ongoing work and measurement, you won't see either happen.
Fix: set a quarterly content refresh cadence. Monitor mention rate weekly. Track competitor visibility. Treat AI visibility as ongoing maintenance, not a one-time project.
15. Single-platform focus (one-size-fits-all)
Each AI platform uses a different source selection system, and their audiences differ in size: Pew's 2026 survey found 44% of U.S. adults use ChatGPT, 24% Gemini and 6% Claude. ChatGPT aligns with Bing, Gemini leverages Google's ecosystem, Perplexity maintains an independent crawl with strong freshness bias, and Claude rewards balanced authority. A tactic that wins on one platform may have zero impact on another.
Fix: track AI visibility per platform separately. Tune content for each platform's preferences -- see Ranqo's platform-specific guide for the breakdown.
What to Fix First
Not every fix has the same effort-to-impact ratio. Start with the "Do first" quadrant -- high-impact fixes that take minutes or hours, not weeks. These four fixes alone can recover much of your missing citation surface.
Fix Priority Matrix
Where to start: high-impact, low-effort fixes first
| Fix | Impact | Effort | Priority |
|---|---|---|---|
| Add author bylines + bios | High | Low | Do first |
| Unblock AI crawlers in robots.txt | High | Low | Do first |
| Add visible "last updated" dates | High | Low | Do first |
| Submit sitemap to Bing Webmaster Tools | High | Low | Do first |
| Implement SSR for content pages | High | High | Plan next |
| Get onto ranked "best X" lists | High | Medium | Plan next |
| Add original statistics to all content | High | Medium | Plan next |
| Open paywalled key content | High | Medium | Plan next |
| Convert prose to tables | Medium | Low | Quick wins |
| Add FAQ schema that mirrors visible Q&A | Medium | Low | Quick wins |
| Expand thin key pages to answer the whole question | Medium | Medium | Quick wins |
| Earn third-party reviews / press | High | High | Long-term |
| Set up monthly content refresh cadence | Medium | High | Long-term |
| Per-platform optimization strategy | Medium | High | Long-term |
For a comprehensive 7-step optimization framework that addresses most of these mistakes systematically, see Ranqo's content optimization playbook.
The 15-Question Self-Audit
Run this audit on your site today. Each "no" is an active anti-GEO mistake costing you AI citations right now.
The 15-Question Self-Audit
If you answer "no" to any of these, you have an active anti-GEO mistake
- 1
Does my site work without JavaScript? (View page with JS disabled)
Maps to mistake #1
- 2
Is my key content (pricing, comparisons, docs) behind a paywall or login?
Maps to mistake #2
- 3
Does my robots.txt allow GPTBot, ClaudeBot, and PerplexityBot?
Maps to mistake #3
- 4
Does every content page have a named author with a bio?
Maps to mistake #4
- 5
Am I tracking my AI mention rate (or only Google rankings)?
Maps to mistake #5
- 6
Do my key informational pages answer the whole question?
Maps to mistake #6
- 7
Do my top 10 pages have FAQ, Article, or Product schema (JSON-LD)?
Maps to mistake #7
- 8
Have my high-value pages been updated in the last 12 months?
Maps to mistake #8
- 9
Does each page have at least 3 original statistics or data points?
Maps to mistake #9
- 10
Have I submitted my sitemap to Bing Webmaster Tools?
Maps to mistake #10
- 11
Do I have dedicated comparison pages for my main competitors?
Maps to mistake #11
- 12
Are my comparisons in HTML tables (not just paragraphs)?
Maps to mistake #12
- 13
Am I listed on G2, Capterra, TrustPilot, or industry review sites?
Maps to mistake #13
- 14
Do I have a quarterly content refresh cadence?
Maps to mistake #14
- 15
Am I tracking visibility per platform (not just one)?
Maps to mistake #15
The fastest path to AI visibility isn't doing more -- it's stopping the things that actively hurt you. Audit first. Optimize second.
Get an instant anti-GEO audit
Run a 6-dimension AI readiness scan on any page: crawlability, content quality, page speed, AI extractability, citation potential, and authority. Each mistake from this list maps directly to a dimension your audit will surface. For the full framework, see the audit guide and 5 factors that drive AI citations.
Start your auditWritten by
Nisha Kumari
Nisha Kumari is Co-Founder at Ranqo, where she leads growth strategy and client acquisition. With a background in digital marketing and financial management, she specializes in SEO, Generative Engine Optimization, and helping brands build visibility across AI platforms.
Share this article
Related articles
The 10 Best AI Citation Tracking Tools (2026)
Only 2.9% of AI citations point to a brand's own site, so source tracking is mostly about other people's pages. Every tool here lists the domains and pages AI cites. What sets them apart is whether they label each source by type and each page by format, then turn that into a next step. Here are 10 tools, ranked on 10 checks.
10 Checks to Find Out Why Your Competitor Shows Up in AI Answers (and You Don't)
In our study, a brand's own website supplied 2.9% of the citations behind AI answers, so the page that put your competitor there is usually someone else's. But one answer is a weak signal, and each engine names different rivals. Check that the gap is real first. Here are 10 checks in the order to run them, each with a free test.
The AI Page Refresh Checklist: Make a Page Citable Without Inventing Facts
The best-known GEO study reports that adding statistics, quotes and sources lifts a page's visibility in AI answers by up to 40%. Its published prompts let the model invent all three. The lesson for a real page is not to add numbers but to add true ones. Here is a 12-point checklist for refreshing a page without inventing anything.