You cannot buy a spot in ChatGPT answers, but you can make your brand the easiest, best-corroborated source to cite. This 10-step LLM SEO playbook covers content formats, schema, llms.txt, Reddit and third-party mentions, and how to track results.
Updated September 2026.
To rank in ChatGPT, make your brand the easiest and best-corroborated source for the questions your buyers ask. In practice that means: let AI crawlers access your site, publish pages that answer specific prompts directly with citable facts and sources, add schema markup, and earn mentions on the third-party sites ChatGPT already trusts, such as review sites, comparison listicles, and Reddit threads.
There is no paid placement and no guaranteed position. ChatGPT, Perplexity, Gemini, and Google AI Overviews generate answers from sources they retrieve and trust. This discipline is called LLM SEO, ChatGPT SEO, or more formally generative engine optimization (GEO).
If you do not have time to run this playbook every week, Dooza Ranky, Dooza's AI Visibility & Growth Employee, automates most of it from $49/mo.
ChatGPT answers come from two places:
Retrieval is where most businesses can make fast progress. For a given prompt, the model tends to favor sources that are relevant to the exact question, easy to extract a clean answer from, specific and verifiable, and in agreement with other sources.
The research backs this up. The Princeton-led paper "GEO: Generative Engine Optimization" tested content changes on a benchmark of queries and found that adding citations to sources, quotations, and statistics boosted visibility in generative engine responses by up to roughly 40%, while traditional keyword stuffing did little.
The fundamentals are shared, but each engine has quirks worth knowing:
| Answer engine | How it sources answers | Extra priority |
|---|---|---|
| ChatGPT | Training data plus live web search with citations | Allow OAI-SearchBot; be present on widely cited third-party pages |
| Perplexity | Retrieval-first; almost every answer shows numbered sources | Allow PerplexityBot; fresh, well-structured pages; Reddit and forum presence |
| Google AI Overviews / AI Mode | Built on Google Search's index and ranking systems | Classic SEO strength; Google says no special markup is required (source) |
| Gemini | Model knowledge plus Google Search grounding | Strong Google presence, Google Business Profile for local |
| Claude | Model knowledge plus web search when enabled | Allow Claude-SearchBot and Claude-User |
| Microsoft Copilot | Grounded in Bing | Submit your site to Bing Webmaster Tools |
This is why GEO and SEO are complements: strong search rankings feed several answer engines at once.
Write down 20–50 questions real buyers ask. Pull them from sales calls, support tickets, and Google's "People also ask." Include:
Enterprise platforms such as Profound estimate these with Prompt Volumes. Small teams can start with a spreadsheet.
Ask each prompt in ChatGPT, Perplexity, Gemini, and Google (check the AI Overview). Record whether you are mentioned, your position, the sentiment, and which URLs are cited. The cited URLs are your target list: they are the pages the engines already trust.
Check robots.txt and your CDN or firewall settings. Allow Googlebot, Bingbot, OAI-SearchBot, ChatGPT-User, PerplexityBot, Claude-SearchBot, and Claude-User on public pages. Make sure key content is server-rendered HTML, not hidden behind JavaScript-only widgets.
Each page should open with a direct answer in the first 100–150 words, then expand. Use the prompt as the H1 or an H2. One page can cover a cluster of closely related prompts.
Follow the Princeton findings. Add specific numbers with sources, short quotations from credible people, and outbound links to primary sources. Replace "affordable pricing" with "plans start at $49/mo." Specific claims give models something concrete to repeat.
Use question-style H2s, comparison tables, numbered steps, short paragraphs, and an FAQ block. Tables are especially useful for "vs," "best," and pricing prompts.
Implement Organization (with logo, sameAs links to your social profiles), Product or Service, Article with author and dates, FAQPage, and LocalBusiness if relevant. Schema makes entities and facts unambiguous.
llms.txt is a proposed standard: a Markdown file at your site root that summarizes your business and links to your most important pages. It is not a ranking factor any engine has confirmed, but it is cheap to add and gives AI tools a clean map of your site.
This is the step most brands skip, and often the one that matters most. Answer engines reflect consensus. Work through the cited-URL list from Step 2:
Update priority pages regularly with current facts and dates. Re-run your prompt set monthly, compare against the baseline, and move effort toward prompts where you are still absent.
Use this skeleton for every priority prompt. It works for ChatGPT, Perplexity, Gemini, and Google at the same time:
A useful self-test: copy your first paragraph into a message on its own. If it fully answers the prompt without the rest of the page, you have written something an answer engine can cite.
| Format | Why answer engines use it | Example |
|---|---|---|
| Comparison pages ("X vs Y") | Structured differences are easy to summarize | Dooza vs Profound |
| Alternatives lists | Match "alternatives to X" prompts directly | Profound alternatives |
| "Best X for Y" roundups | Match recommendation prompts | Best GEO tools |
| Pricing explainers | Concrete numbers are highly citable | Profound AI pricing |
| Definitions and guides | Supply the "what is" sentence models quote | What is GEO? |
| Original data and surveys | Unique statistics get cited as the source | Your own customer benchmarks |
| FAQ blocks | One question, one extractable answer | End of every strategic page |
Track four things monthly:
You can do this manually in a spreadsheet for 20–30 prompts. For scale, tools such as Profound (Answer Engine Insights and Agent Analytics), Peec AI, and Otterly automate it. Our GEO tools comparison covers the options and prices.
There is no fixed timeline, and nobody can guarantee a mention. Retrieval-based engines like Perplexity and ChatGPT search can pick up a well-structured new page once it is crawled and indexed. Earning third-party mentions and shifting what a model "knows" about your category takes longer, typically months of consistent work. Plan for a quarter of steady execution before judging results.
Ranky is Dooza's AI Visibility & Growth Employee. It runs most of the playbook for you, with approvals where you want them:
| Playbook step | What Ranky does |
|---|---|
| Prompt and topic research | Researches keywords and topics with real search data |
| Answer pages and citable content | Writes GEO-ready articles with direct answers, sourced claims, tables, FAQs; embeds YouTube videos and Reddit threads |
| Publishing cadence | Publishes daily or 3x a week to WordPress, Shopify, Wix, or custom sites |
| Schema and on-page SEO | Handles titles, meta, schema markup, and internal links |
| Third-party presence | Listens on Reddit, LinkedIn, and YouTube and drafts brand-voice comments to approve or auto-send |
| Consistency | Keeps name, address, and phone consistent for local answers |
| Reporting | Monitors who mentions you and sends a nightly recap |
Ranky is an execution employee, not an enterprise analytics suite, so teams that want multi-engine dashboards often pair it with a tracker. See what is included in our generative engine optimization service.
You now have the playbook. The hard part is running it every week. Book a call with a Dooza engineer and we will set up your first AI employee free, live in days, with no contract. Or start Ranky on Dooza Workforce from $49/mo.
Allow AI crawlers, publish pages that answer buyer prompts directly with citable facts and sources, add schema markup, and earn mentions on third-party sites ChatGPT already cites, such as review sites, listicles, and Reddit.
No. There is no paid placement in ChatGPT's organic answers. Visibility comes from being a relevant, trustworthy, and well-corroborated source.
LLM SEO is the practice of optimizing your content and brand presence so large language models like ChatGPT, Claude, and Gemini mention and cite you. It is another name for generative engine optimization (GEO).
Be described consistently across trusted third-party sources, publish specific and verifiable facts about your brand, and make sure OAI-SearchBot can crawl your site so ChatGPT search can retrieve and cite your pages.
llms.txt is a proposed standard that gives AI tools a clean map of your site. No major engine has confirmed it as a ranking factor, but it is low-effort and can help AI tools find key pages.
Perplexity is retrieval-first and shows sources for nearly every answer, so fresh, well-structured, crawlable pages and forum presence matter a lot. ChatGPT also draws on model knowledge from training data.
It varies. Retrieval-based answers can pick up a new indexed page relatively quickly, while changing what models know about your category usually takes months of consistent content and mentions. No one can guarantee placement.
Yes. Dooza Ranky researches topics, writes and publishes citable content, handles schema and on-page SEO, and drafts community comments, from $49 per month with free engineer setup for your first AI employee.
Automate your business with AI employees that work 24/7.
SEO earns rankings in search results. GEO earns mentions and citations inside AI answers from ChatGPT, Perplexity, Gemini, and Google AI Overviews. Here is how they differ, where they overlap, and how to do both without doubling your workload.
Generative engine optimization (GEO) is how brands get mentioned and cited in AI answers from ChatGPT, Perplexity, Gemini, and Google AI Overviews. This guide explains how answer engines choose sources, the tactics that work, how to measure AI search visibility, and the tools and services available.
Join thousands of companies using Workforce to automate their work. Get started for free today.
No credit card required · 7-day money-back guarantee · Cancel anytime