How ChatGPT, Gemini, Google AI Mode, Perplexity and Copilot choose sources, and how iGaming brands get cited accurately.
AEO (answer engine optimisation) is the work of getting your brand cited and described accurately in AI answers from ChatGPT, Gemini, Google AI Overviews and AI Mode, Perplexity and Copilot. These systems run web searches, mostly against Google’s and Bing’s indexes or their own crawlers, and quote pages that answer the question directly. For iGaming brands that means four things: let the crawlers in (geo-blocks and age gates often keep them out), answer buyer and player questions in two or three plain sentences, keep facts consistent across your site, directories and press, and get covered on the comparison pages and articles the engines already cite.
On this page
- What is AEO and how is it different from SEO?
- How do ChatGPT, Gemini and Google AI Mode choose sources?
- How do citations work, and why do some pages get cited more?
- Which questions should an iGaming brand target?
- What kind of content gets cited?
- Does an iGaming site need llms.txt?
- Which schema helps AI search?
- Why does Bing indexing matter for AI search?
- What blocks iGaming sites from being cited?
- What limits apply to gambling brands in AI answers?
- How do you measure AI search visibility?
- Which regulator sources do assistants use to check licences?
- Worked example: building a prompt set
- Step by step: a 30-day AEO plan
- What are the common AEO mistakes?
- AEO checklist
- What does AEO work with Entri involve?
- Further reading
- Frequently asked questions
For iGaming, AEO matters in two places. Players ask assistants which casino or sportsbook to trust, how long withdrawals take and whether a brand is licensed. Operator buying teams ask which platform, payment provider or support vendor to shortlist. In both cases the assistant gives a short answer with a handful of sources, and if you are not one of them you are not in the conversation.
What is AEO and how is it different from SEO?
SEO aims to rank a page in a list of results. AEO aims to be one of the few sources an AI system quotes when it writes an answer. The two overlap heavily, because most AI search features retrieve pages from a search index before writing. The differences are in emphasis:
- The answer matters more than the page. A two-sentence direct answer high on the page is easier to quote than a long introduction.
- Third-party mentions carry more weight. Engines often cite comparison pages, listicles, directories and news rather than vendor homepages.
- Consistency matters. If your site, your directory profiles and press coverage give different facts, the assistant may pick the wrong one or hedge.
- There is no position one. You are either cited or not, and the answer changes with the wording of the question, the market and the language.
How do ChatGPT, Gemini and Google AI Mode choose sources?
Each system works differently, and each publishes some guidance.
| Engine | Where it finds pages | Crawler or control | What helps most |
|---|---|---|---|
| Google AI Overviews and AI Mode | Google’s search index, with “query fan-out” across sub-questions | Googlebot; snippet controls such as nosnippet | Indexed, snippet-eligible pages that answer sub-questions directly |
| Gemini app | Google Search grounding for current information | Googlebot for Search; Google-Extended for use in Gemini models | Same as Google Search, plus clear entity facts |
| ChatGPT search | OpenAI’s own crawler and search partners | OAI-SearchBot (search), GPTBot (training), ChatGPT-User (user requests) | Allowing OAI-SearchBot; comparison pages and direct answers |
| Perplexity | Its own index plus live fetching | PerplexityBot and Perplexity-User | Fresh, well-structured pages with sources |
| Microsoft Copilot | Bing’s index | Bingbot; IndexNow for fast updates | Bing indexing and Bing Webmaster Tools |
Google AI Overviews and AI Mode. Google’s documentation on AI features and your website says there are “no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary”. A page must be indexed and eligible to show with a snippet. Both features may use “query fan-out”, running several related searches across subtopics to build a response, which is why they often cite pages that rank for a sub-question rather than the main query. Clicks and impressions from these features are reported in Search Console’s Performance report under the Web search type.
Gemini. The Gemini app uses Google Search to ground answers about current topics, so the Google Search basics apply. Google’s documentation also describes Google-Extended, a robots.txt token that controls whether content is used for training and grounding in Gemini models. Blocking it does not affect Google Search.
ChatGPT. OpenAI’s crawler documentation separates OAI-SearchBot, which decides whether a site can appear in ChatGPT search results, from GPTBot, which collects training data, and ChatGPT-User, which fetches pages when a user asks. Blocking GPTBot does not remove you from ChatGPT search; blocking OAI-SearchBot does.
Perplexity. Perplexity documents PerplexityBot and its user-triggered agent in its bots guide.
Copilot and Bing. Microsoft Copilot answers from Bing, so Bing indexing matters as much as Google’s. See the section on Bing below.
How do citations work, and why do some pages get cited more?
An AI answer is usually built in three steps: the engine rewrites the question into one or more searches, retrieves a set of pages, then writes an answer and attaches links to the pages it relied on. Pages get cited more when:
- They match a sub-question exactly. A heading such as “How long do withdrawals take at [brand]?” followed by a direct answer matches the fan-out query.
- The answer is self-contained. A paragraph that makes sense when lifted out of the page, with the subject named rather than “it” or “we”.
- They contain specific, sourced facts. Numbers, dates, licence numbers and named sources are easier for an engine to trust and quote.
- They compare options. For “best X for Y” questions, engines lean heavily on tables with clear criteria.
- Other sites say the same thing. Agreement between your site, directories, trade press and partner pages raises confidence.
Which questions should an iGaming brand target?
Build a prompt set: the 30 to 100 questions your players or buyers actually ask, grouped by intent. Search Console queries, sales call notes, support tickets and “People also ask” boxes are good sources.
| Audience | Example questions |
|---|---|
| Players | Is [brand] licensed in Ontario? How long do [brand] withdrawals take? Which UK casinos pay out fastest? |
| Operator buyers | What is the best AI customer support for online casinos? Which iGaming platforms support Brazil? Which tools give analytics for retention and deposits? |
| Affiliates and partners | Which operators offer revenue share in Brazil? What does [operator] pay affiliates? |
| Press and analysts | How big is the GB online casino market? What changed in UK bonus rules in 2026? |
Run each question in ChatGPT, Gemini, Google AI Mode, Perplexity and Copilot, in each language and market that matters. Record whether you are mentioned, how you are described, and which URLs are cited. That is your baseline.
What kind of content gets cited?
- Direct answers under question headings. Write the buyer’s question as the heading and answer it in two or three sentences before any background.
- Comparison tables and listicles. If you sell to operators, publish honest comparisons in which you describe competitors fairly.
- An official facts page. One page listing what you are, what you sell, where you operate, certifications and key numbers. Cevro’s About Cevro for AI and LLMs page is one example of this format from an iGaming supplier.
- Help centre answers. For player questions about withdrawals, verification and bonuses, public help pages are often the best source an engine can find. Sister site igamingsupportai.com explains how support content helps SEO and AI search.
- Third-party coverage. Directory profiles, trade press, podcasts and course material. A profile on a B2B iGaming directory, a guest spot on an iGaming and AI podcast or a mention in course material all give engines more independent sources to cite.
Does an iGaming site need llms.txt?
llms.txt is a proposed convention, described at llmstxt.org, for a Markdown file at the root of a site that summarises what the site is and links to its most useful pages, written for language models. It is not an official standard, and Google says no new machine-readable files or AI text files are needed to appear in its AI features.
Our view: it is cheap to publish and does no harm, so add one, but do not expect it to move citations on its own. Keep it short: a one-paragraph summary, your main service or product pages, your facts page and your key guides, each with a one-line description. This site publishes one at entri.io/llms.txt.
Which schema helps AI search?
Structured data (schema.org markup) helps search engines understand what a page and a business are. Useful types for iGaming sites:
- Organization with legal name, logo, official profiles in sameAs, and contact points. See Google’s Organization markup guide.
- WebSite for the site name.
- Article for guides and news, with author and dates.
- FAQPage for question and answer sections. Since August 2023 Google shows FAQ rich results mainly for authoritative government and health sites (Google Search Central), but the markup still describes questions and answers clearly.
- Product or SoftwareApplication for B2B suppliers, where accurate.
Google is clear that markup must match what users can see on the page. Markup is a supporting signal, not a shortcut.
Why does Bing indexing matter for AI search?
Bing’s index feeds Microsoft Copilot, and Bing results also appear in partner search engines. Many iGaming sites check Google Search Console and never open Bing Webmaster Tools. Basic steps:
- Verify the site in Bing Webmaster Tools and submit the sitemap.
- Check that key pages are indexed and that Bingbot is not blocked by firewall or geo rules.
- Use IndexNow, an open protocol supported by Bing and other engines, to notify them when pages are published or updated. On WordPress, Microsoft’s free IndexNow plugin does this automatically.
- Review Bing’s webmaster guidelines.
What blocks iGaming sites from being cited?
Gambling sites have crawl problems most industries do not:
- Geo-blocking. Licensed operators block visitors from markets where they are not licensed. Many crawlers fetch from the United States, so a UK or Ontario operator that blocks US IP addresses may block Googlebot, Bingbot or OAI-SearchBot from its help and information pages. Verify crawlers properly (Google explains how to verify Googlebot) and let them read public information pages.
- Age gates and login walls. Put age verification in front of games, not in front of help pages, licence pages and guides.
- Cookie and consent walls that hide content until a click.
- CDN and firewall bot rules. Security tools can block AI crawlers by default. Check your rules for each named bot.
- JavaScript-only content. Answers that appear only after scripts run may not be read.
- Robots.txt. Check that Googlebot, Bingbot, OAI-SearchBot and PerplexityBot are allowed, and decide deliberately about GPTBot and Google-Extended.
What limits apply to gambling brands in AI answers?
AI assistants are cautious with gambling. They may decline to recommend operators in some markets, add responsible gambling warnings, or prefer regulator and news sources. That makes accurate, compliant content more important:
- State licence details and the markets you are licensed in.
- Do not publish claims you cannot prove. Assistants repeat them, and regulators read them.
- Keep bonus terms current, since an assistant quoting an expired offer creates complaints.
- For B2B suppliers, publish verifiable proof such as certifications and named customer stories approved by the customer.
How do you measure AI search visibility?
- Citation share. For each question in the prompt set, are you cited, mentioned without a link, or absent? Track monthly.
- Description accuracy. Is what the assistant says about you correct and current?
- Search Console and Bing Webmaster Tools. Watch the pages that answer prompt-set questions.
- Referral traffic. Analytics shows visits from chatgpt.com, perplexity.ai, copilot.microsoft.com and gemini.google.com.
- Assisted conversions. Add “AI assistant” as an option in “how did you hear about us” on sign-up and contact forms.
Which regulator sources do assistants use to check licences?
When a player asks an assistant whether a brand is licensed, the best answers cite the regulator. Make sure your licence pages match these sources exactly and link to them.
| Market | Public source that lists licensed or registered operators | What to do on your own site |
|---|---|---|
| Great Britain | Gambling Commission public register | State the legal entity and account number exactly as registered |
| Ontario | iGaming Ontario regulated market list | Make sure the listed domain is the one you promote and link to the list |
| Brazil | Secretariat of Prizes and Bets | Show company name, CNPJ and authorisation number, as ads must since July 2026 |
Worked example: building a prompt set
The brands below are hypothetical, to show the method.
An Ontario sportsbook (“BrandY”). The team writes 40 prompts across four groups: 12 trust questions (“Is BrandY legal in Ontario?”, “Is BrandY safe?”), 10 payment questions (“How long do BrandY withdrawals take?”, “Does BrandY accept Interac?”), 10 product questions (“Does BrandY have same game parlays?”) and 8 category questions (“Best sportsbook apps in Ontario”). They run all 40 in ChatGPT, Gemini, Google AI Mode, Perplexity and Copilot.
Baseline: BrandY is cited in 9 of 40 answers in ChatGPT and 14 of 40 in AI Mode. Most misses are payment questions, answered from a forum thread and an affiliate page with an out-of-date withdrawal time. The fix is a public payments help page with a two-sentence answer per method, plus a corrected affiliate page. Two months later payment questions cite BrandY’s help page in most engines.
A B2B CRM supplier (“VendorZ”). Thirty prompts such as “best CRM for online casinos”, “VendorZ vs [competitor]” and “which tools give analytics for retention and deposits in iGaming”. Baseline: named in 4 of 30 answers, described with a product it retired. Fixes: an official facts page, a fair comparison page, updated directory profiles and a guest article on a trade site that engines already cite.
Step by step: a 30-day AEO plan
- Week 1: build the prompt set and run the baseline across engines, markets and languages.
- Week 1: crawl audit: robots.txt, geo-blocking, age gates, CDN bot rules, Search Console and Bing Webmaster Tools coverage.
- Week 2: fix access problems and submit sitemaps to Google and Bing; turn on IndexNow.
- Week 2 to 3: publish or rewrite answer pages for the questions you lose most, plus an official facts page.
- Week 3: add or correct schema and publish llms.txt.
- Week 4: off-site corrections and outreach to the sources that win your questions.
- Week 4: re-run the prompt set and set the monthly tracking routine.
What are the common AEO mistakes?
- Blocking crawlers by accident through geo-blocking, age gates or CDN defaults.
- Blocking OAI-SearchBot when you meant to block GPTBot, which removes you from ChatGPT search.
- Writing long introductions before the answer.
- Inconsistent facts across your site, directories and press.
- Publishing dozens of near-identical “best” pages, which can look like scaled content abuse.
- Expecting llms.txt or schema alone to fix visibility.
- Tracking only Google. Copilot runs on Bing, and ChatGPT and Perplexity have their own crawlers.
AEO checklist
- Prompt set of 30 to 100 questions per audience and market
- Baseline citation share recorded per engine
- Googlebot, Bingbot, OAI-SearchBot and PerplexityBot allowed and not geo-blocked
- Help, licence and guide pages readable without age gate or login
- Site verified in Search Console and Bing Webmaster Tools, sitemaps submitted
- IndexNow enabled
- Official facts page published and linked from the footer
- Organization, WebSite, Article and FAQPage schema matching visible content
- llms.txt published
- Monthly tracking of citations and description accuracy
What does AEO work with Entri involve?
- Prompt set and baseline across engines, markets and languages.
- Crawl audit covering robots.txt, geo-blocking, age gates, CDN bot rules, Bing indexing and IndexNow.
- Gap analysis. Which questions you lose, which sources win and why.
- Content and technical fixes. Answer pages, comparison pages, facts page, schema, llms.txt.
- Off-site work. Directory profiles, trade press, partner pages and corrections to outdated third-party content.
- Monthly tracking of citation share and accuracy.
Want to know what AI assistants say about your brand?
Send us your brand, your markets and the questions your buyers ask. We start with a citation baseline.
Further reading
- Brand protection SEO for iGaming
- Content marketing for iGaming brands
- Support MCP explained on igamingsupportai.com
- Cevro Scribe: ask your support data questions from Claude or ChatGPT
- Google: creating helpful, reliable, people-first content
Frequently asked questions
What is AEO?
AEO stands for answer engine optimisation. It is the work of getting your brand cited and described accurately in AI-generated answers from tools such as ChatGPT, Gemini, Google AI Overviews and AI Mode, Perplexity and Copilot.
How does Google AI Mode choose which sites to cite?
It uses Google’s search index and may run several related searches across subtopics (query fan-out), then cites pages that are indexed, eligible for a snippet and answer those sub-questions well. Google says no special optimisation is required beyond normal SEO.
Which bot controls whether my site appears in ChatGPT search?
OAI-SearchBot. Blocking GPTBot stops your content being used for training but does not remove you from ChatGPT search. Check that your robots.txt and firewall allow OAI-SearchBot.
Do I need an llms.txt file?
It is optional. llms.txt is a proposed convention, not a standard, and Google says no AI text files are needed for its AI features. It is cheap to publish, so many sites add a short one.
Does schema markup help AI search?
It helps search engines understand your organisation and pages, which supports accurate answers. Use Organization, WebSite, Article and FAQPage markup that matches visible content. It is a supporting signal, not a ranking shortcut.
Why does Bing matter if most players use Google?
Microsoft Copilot answers from Bing’s index, and Bing results also appear in partner search engines. Verifying the site in Bing Webmaster Tools and using IndexNow keeps it current.
Why might an iGaming site be invisible to AI crawlers?
Geo-blocking of the countries crawlers use, age gates and login walls in front of information pages, consent walls, CDN bot blocking and robots.txt rules are the common causes.
How long does AEO take?
Technical fixes and new answer pages can be cited within weeks once indexed. Changing how assistants describe a brand across many questions usually takes several months of content and off-site work.
What is a prompt set?
A fixed list of 30 to 100 questions your players or buyers ask, grouped by intent and market, that you run in each AI assistant every month to measure whether and how you are cited.
Can geo-blocking stop my site appearing in AI answers?
Yes. Many crawlers fetch from the United States, so blocking US traffic can block Googlebot, Bingbot or OAI-SearchBot. Verify crawlers and let them read public information pages.