ChatGPT, Perplexity and Google's AI Overviews answer a question in one block of text and cite only a few pages. The researchers who introduced generative engine optimization (GEO) reported that their methods raised visibility in those answers by up to 40% on a benchmark of varied queries (KDD 2024 paper). This GEO checklist shows you what to fix first. Kwebby's interactive version has 21 items in five groups, from AI crawler access to measuring which pages assistants cite.
Key Takeaways
- Kwebby's GEO Checklist has 21 items in five groups: AI Crawler Access (5), Answer-Ready Content (5), Facts, Sources & Trust (4), Entities & Structured Data (4) and Measurement & Optional Files (3).
- Six items are High priority, fourteen are Medium and one is Low, so you know where to start.
- Google says no extra requirements or special optimizations apply to AI Overviews and AI Mode. A page must be indexed and eligible to appear with a snippet.
- AI crawlers have different jobs. OpenAI says blocking GPTBot, its training crawler, does not affect how a site appears in ChatGPT search, which uses OAI-SearchBot.
- llms.txt is a proposal with limited adoption. Google says you do not need AI text files to appear in its AI features.
- Your ticks are saved in your browser, so you can re-run the audit after each release. A full score does not promise citations.
What Is the Kwebby GEO Checklist?
Kwebby's GEO Checklist is a free interactive page with 21 audit items that you tick off while a percentage score updates. GEO, short for generative engine optimization, means preparing your content so AI assistants can find it, trust it and cite it in generated answers.
The page shows a Score, a Complete counter that starts at 0 / 21, and a count of High (6), Medium (14) and Low (1) items. Each item carries a one-line instruction, and many link to the free Kwebby tool that helps.
| Group | Items | High | Medium | Low |
|---|---|---|---|---|
| AI Crawler Access | 5 | 3 | 2 | 0 |
| Answer-Ready Content | 5 | 1 | 4 | 0 |
| Facts, Sources & Trust | 4 | 1 | 3 | 0 |
| Entities & Structured Data | 4 | 1 | 3 | 0 |
| Measurement & Optional Files | 3 | 0 | 2 | 1 |
| Total | 21 | 6 | 14 | 1 |
GEO builds on SEO instead of replacing it. For deeper coverage of one area, open Kwebby's SEO Checklist or Technical SEO Checklist. You can also browse every list on the SEO checklists hub.
How GEO Relates to SEO
GEO and SEO share the same foundation: a page that search engines can crawl, index and understand. Google states that its AI features need no additional requirements or special optimization, and that a page must be indexed and eligible to show with a snippet (Google Search Central).
The goal differs. SEO earns a click from a ranked list. GEO aims to get your page quoted or cited inside a generated answer, where the reader may never see a list at all. That is why the checklist focuses on crawler access, quotable passages and verifiable facts.
How to Use the Interactive GEO Checklist
Open the page, tick every item your page already passes, and fix the unticked high-priority items first. Audit one URL at a time, such as your homepage or a key guide.
- Open the checklist. The Score starts at 0 percent and Complete at 0 / 21.
- Read the priority tag. Each item shows high, medium or low, plus a one-line instruction.
- Tick what passes. Each tick raises the Score and the Complete counter, and your ticks are saved in this browser.
- Fix the high items first. Six items carry the high tag, from indexing to named sources.
- Measure. Ask AI assistants your key questions and log which pages they cite.
- Repeat after every release. Press Reset to start a fresh audit of the same page or a new one.
The Five Checklist Groups Explained
Each group answers a question an AI system asks about a page: can I reach it, can I quote it, can I trust it, do I know who it is about, and is it working?
AI Crawler Access (5 items)
Three of the five items are high priority, because a blocked or empty page cannot be cited.
- Confirm the page is indexed in Google and Bing (High). The Google Index Checker gives a quick view of what Google has indexed.
- Set robots.txt rules per AI crawler purpose (High). Block by purpose, not by brand. The table below lists the main tokens. The Robots.txt Generator switches 36 search and AI bots on or off.
- Check CDN and firewall bot settings (Medium). Some CDNs and firewalls offer a single setting that blocks AI crawlers. Make sure it matches what your robots.txt allows.
- Serve the main content in the HTML (High). Not every AI crawler runs JavaScript. Open the raw HTML in the Source Code Viewer and confirm the main text, headings and links are there.
- Keep the XML sitemap and lastmod accurate (Medium). The XML Sitemap Generator builds a sitemap.xml file.
| Token | Owner | Purpose |
|---|---|---|
| OAI-SearchBot | OpenAI | Surfaces sites in ChatGPT search results |
| GPTBot | OpenAI | Collects data for training models |
| Claude-SearchBot | Anthropic | Improves Claude's search results |
| ClaudeBot | Anthropic | Collects data for training models |
| PerplexityBot | Perplexity | Surfaces and links sites in Perplexity answers, not used for training |
| Google-Extended | A robots.txt token, not a separate crawler. Controls use in Gemini training and grounding. It does not affect Google Search |
Sources: OpenAI, Anthropic, Perplexity and Google. OpenAI says a robots.txt change can take about 24 hours to take effect. User-triggered agents such as ChatGPT-User, Claude-User and Perplexity-User fetch a page when a person asks. OpenAI and Perplexity say robots.txt rules may not apply to their user agents in the same way, while Anthropic lists Claude-User among the bots that follow robots.txt.
Answer-Ready Content (5 items)
This group makes your passages easy to quote. One item is high priority.
- Lead each section with a direct answer (High). Answer the heading's question in the first one or two sentences.
- Use headings that match real questions (Medium). Phrase H2 and H3 headings the way your customers ask.
- Use lists and tables for steps and comparisons (Medium). Structured formats are easy to scan and to extract.
- Write self-contained passages (Medium). Name the entity in full, such as "Kwebby's Robots.txt Generator", instead of writing "it".
- Cover the follow-up questions (Medium). Google describes AI Mode as running several related searches for one question, so answer the next questions a reader would ask.
Facts, Sources & Trust (4 items)
This group gives an AI system a reason to trust and cite you.
- Back every claim with a named source (High). Link to the primary source for statistics, rules and quotes. Never publish invented numbers.
- Add original data, quotes or examples (Medium). In the GEO paper, adding quotations lifted visibility 41%, statistics 31% and cited sources 27% on its main metric, while keyword stuffing gave little to no improvement.
- Show the author and their credentials (Medium). A byline, short bio and author page show who stands behind the facts.
- Update facts and show the date (Medium). Stale facts are the first thing a reader or an AI summary can spot.
Entities & Structured Data (4 items)
This group helps systems connect your pages to the right business, product or person.
- Keep entity facts consistent everywhere (High). Use the same business name, address, product facts and author names on your site and on your profiles.
- Add Organization markup with sameAs links (Medium). The Schema Markup Generator builds Organization and Person JSON-LD.
- Match structured data to visible content (Medium). Google's rules say structured data must describe content visible on the page (Google). Google's changelog also shows the FAQ rich result as deprecated in 2026. FAQPage markup stays valid schema.org, but do not expect a Google FAQ display from it.
- Link related pages with descriptive anchors (Medium). The Website Links Count Checker lists the internal links on any page.
Measurement & Optional Files (3 items)
This group tells you whether the work pays off.
- Track AI referrals in analytics (Medium). Segment visits from chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com. Search Console counts AI Overviews and AI Mode within its web search data, without a separate filter.
- Test your key questions in AI assistants monthly (Medium). List ten questions your buyers ask, put them to ChatGPT, Perplexity, Gemini and Google AI Mode, and log whether your page is cited.
- Treat llms.txt as optional (Low). llms.txt is a 2024 proposal for a Markdown file that points language models to key pages. Adoption by AI providers is limited, and Google says you do not need new machine-readable files or AI text files to appear in AI Overviews or AI Mode. Add one only if it costs you little.
Mistakes That Waste a GEO Effort
- Blocking everything. A broad
Disallow: /rule removes you from classic search and AI search at once. Test your robots.txt after every edit. - Blocking search bots to stop training. GPTBot and ClaudeBot handle training. Blocking OAI-SearchBot or Claude-SearchBot removes you from those search features.
- Treating the score as a guarantee. The checklist measures preparation, not placement.
- Repeating keywords. Google's spam policies name keyword stuffing, and the GEO paper found it brought little to no gain.
- Marking up what readers cannot see. Hidden or invented markup breaks Google's guidelines.
- Hiding content behind JavaScript. A crawler that does not render scripts sees an empty page.
Frequently Asked Questions (FAQs)
What does GEO stand for?
GEO stands for generative engine optimization. It is the practice of preparing content so AI assistants can find it and cite it in generated answers. The term comes from a 2023 research paper accepted to KDD 2024.
Is GEO different from SEO?
GEO overlaps heavily with SEO. Both need crawlable, indexed, well-structured pages. GEO adds a focus on being quoted inside a generated answer, through answer-first passages, verifiable facts and crawler access for AI search bots.
Do I need an llms.txt file to appear in AI answers?
No. llms.txt is a proposal with limited adoption. Google states that appearing in AI Overviews and AI Mode needs no new machine-readable files, AI text files or special markup. The page must be indexed and eligible to show with a snippet.
Should I block AI crawlers in robots.txt?
It depends on the crawler's purpose. GPTBot and ClaudeBot collect training data, while OAI-SearchBot, Claude-SearchBot and PerplexityBot power search features. Google-Extended controls Gemini training and grounding and does not affect Google Search. OpenAI and Perplexity say their user-triggered agents may not follow robots.txt rules in the same way.
How many items are on the Kwebby GEO Checklist?
The checklist has 21 items in five groups. Six are high priority, fourteen are medium and one is low. A Score and a Complete counter update as you tick each item.
Does the checklist save my progress?
Yes. Your ticks are saved in your browser's local storage, so they are still there when you return on the same device and browser. Nothing is sent to Kwebby. The Reset button clears them.
Final Thoughts
A GEO checklist works best as a routine, not a one-off audit. Run Kwebby's 21 items on your most important page, fix the high-priority gaps, then measure which pages AI assistants cite. Open the SEO checklists hub if you want a second list for technical or local work. Check your robots.txt first, because a blocked crawler cannot cite a page it cannot read.