More buyers now start research by asking an assistant a question. "What's the best AI code review tool for a small team?" "Which influencer agency works with AI startups?" They get back a few paragraphs and a short list of names. If your brand is in that list, you're in the running. If it isn't, the buyer may never see your site at all.
GEO is the discipline that grew up around that moment. This post covers where the term came from, how these assistants decide what to say, what changes compared with SEO, and what the work looks like in practice.
Where does the term GEO come from?
The phrase "generative engine optimization" was formalized in a 2023 research paper by a team from Princeton University and IIT Delhi, later published at the KDD 2024 conference (Aggarwal et al., "GEO: Generative Engine Optimization"). The authors built a benchmark of 10,000 queries and tested different ways of rewriting web pages to see which ones made a source more visible in AI-generated answers.
Two findings are worth knowing. Adding citations, quotations from credible sources and statistics raised a page's visibility in generated answers by up to about 40% on their benchmark. Keyword stuffing, the old SEO habit of repeating the search phrase, offered little or no improvement, and on Perplexity it did worse than leaving the page alone. The paper also found that lower-ranked pages gained the most from these changes, which is encouraging for smaller brands.
The research is a starting point, and AI products have changed a lot since it was written. The core idea has held up: these systems reward pages that are clear, specific and backed by evidence.
How do AI assistants decide what to say about a brand?
There are two main routes into an AI answer.
The first is what the model already learned during training. Large models are trained on huge amounts of public web text. If your brand was written about often and consistently before the training cutoff, the model may "know" you without looking anything up.
The second is live retrieval. When someone asks a question that needs current information, the assistant runs searches, reads a handful of pages and writes an answer grounded in those sources, usually with links. Google describes AI Overviews and AI Mode as using a "query fan-out" technique, issuing multiple related searches across subtopics before writing a response (Google Search Central).
Each company runs its own crawlers for these jobs. OpenAI, for example, uses GPTBot for training data and OAI-SearchBot for ChatGPT's search features, and says sites that opt out of OAI-SearchBot won't be shown in ChatGPT search answers (OpenAI crawler docs). Perplexity uses PerplexityBot to surface and link websites in its results (Perplexity docs).
So the practical question for any brand is simple to state. Can these crawlers reach your pages, can they read what's on them, and do the pages and the rest of the web say clear things about you?
How is GEO different from SEO?
They overlap more than the hype suggests. Google says plainly that the same SEO best practices apply to AI Overviews and AI Mode, and that there are no extra technical requirements to appear in them. A page has to be indexed and eligible to show with a snippet, and that's it.
What stays the same?
- Crawlers need access. A robots.txt rule or a firewall setting that blocks a bot blocks you from that product.
- Pages need to be indexable, fast and linked from other pages on your site.
- Content needs to be helpful and written for people. Google's guidance on helpful, reliable, people-first content applies here too.
- Structured data should describe what's actually visible on the page.
What changes?
The unit of success changes. In SEO you count rankings and clicks. In GEO you care about whether your brand is mentioned, how it's described, whether your page is cited as a source, and how often you appear compared with competitors for the same prompt.
Rendering matters more. Googlebot renders JavaScript, but many AI crawlers don't. A 2024 analysis by Vercel and MERJ found that none of the major AI crawlers they studied, including OpenAI's crawlers, ClaudeBot and PerplexityBot, executed JavaScript (Vercel). If your site is a single-page app that ships an empty HTML shell, those crawlers may see a blank page.
Third-party sources matter more. An AI answer usually blends several sources. Reviews, comparison articles, Reddit threads, YouTube videos and creator posts can shape how you're described as much as your own site does.
Writing style matters in a specific way. Assistants pull passages they can quote. A page that answers the question in its first two sentences, uses the buyer's own wording and backs claims with sources gives the model something to work with.
| SEO | GEO | |
|---|---|---|
| Goal | Rank in a list of links | Be named and cited inside the answer |
| Main metric | Rankings, clicks | Mentions, citations, share of voice, sentiment |
| Who reads first | Googlebot and Bingbot | Search crawlers plus AI crawlers like OAI-SearchBot and PerplexityBot |
| JavaScript | Google renders it | Many AI crawlers don't |
| Off-site signals | Backlinks | Mentions, reviews and creator content that AI answers cite |
What does GEO work actually involve?
At Dime we split it into three parts, and most serious programs look similar.
An audit. Check AI crawler access, whether key content renders without JavaScript, how content and structured data are organized, and how well your pages fit the prompts buyers actually ask. Then record a baseline of how ChatGPT, Perplexity, Google AI Overviews and Copilot describe you today compared with competitors.
Optimization. Fix technical blockers, restructure pages so they answer questions directly, add structured data and an llms.txt file, and earn the creator and third-party mentions that AI answers draw on.
Tracking. Run the same set of buyer prompts across AI platforms on a schedule. Watch share of voice against competitors, which sources get cited and whether the sentiment is right, then decide what to fix next.
You can read more about how we run each step on our GEO services page.
Do you need an llms.txt file?
llms.txt is a proposed standard for a plain-text file at the root of your site that summarizes what the site is and links to the pages that matter. It's cheap to add and it helps any tool that chooses to read it.
Keep expectations realistic. Google says you don't need new machine-readable files or AI text files to appear in its AI features. Treat llms.txt as a helpful extra on top of a crawlable, well-structured site.
Should you start with SEO or GEO?
Start with the shared foundation, because it pays off in both. Make sure crawlers can reach you, make sure your pages render as real HTML, give every page a clear title and description, and write pages that answer the questions your buyers ask. Then layer on the GEO-specific work: prompt tracking, answer-first rewrites and a plan for third-party mentions.
If you want a fast read on where you stand, our GEO readiness checklist walks through the checks we run first.