---
title: "GEO vs SEO: What Carries Over and What Changes in WordPress"
url: https://hostmy.blog/geo-vs-seo-wordpress/
date: 2026-10-16
modified: 2026-10-16
lang: en
author: "Aditya Sharma"
description: "For Google, AI search is still SEO. For every other assistant, four things change: who can fetch you, what gets quoted, how you measure, and which engines you serve."
categories:
  - "RankReady"
image: https://hostmy.blog/wp-content/uploads/2026/10/hmb-card-2491-1024x538.jpg
word_count: 1567
---

# GEO vs SEO: What Carries Over and What Changes in WordPress

GEO, or generative engine optimisation, is the work of getting your pages used and quoted in AI-generated answers. SEO is the work of getting them found and ranked. On a WordPress site, most of that work is the same work.

GEO vs SEO is a smaller gap than most sales pages suggest, and a larger one than Google's own advice implies. It sits in four places: who is allowed to fetch your pages, what gets quoted, how you measure it, and which engines you are optimising for. This post covers the shared work first, then those four places, so you can see which jobs your SEO plugin already does.

## Google's answer: it is still SEO

Start with the one source that is not selling anything. Google's guide to generative AI in Search says that [optimising for generative AI search is optimising for the search experience, and so is still SEO](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide).

The same page lists what you can ignore for Google: chunking your content, creating AI text files such as `llms.txt`, and chasing inauthentic mentions. It points to the Generative AI performance report in Search Console as the place to watch how your pages show up in AI features. The guide was last updated in July 2026.

For Google's AI Overviews and AI Mode, that settles most of the question. Your titles, headings, internal links, page speed, author details and honest dates already count, and your SEO plugin already handles part of it. The catch is that Google is one engine among several. ChatGPT Search, Perplexity and Claude each fetch the web with their own crawlers, and a rule you wrote for Googlebot does not speak for them.

## What carries over, task by task

Here is the shared work, and where each job lives on a WordPress site. Nothing in this table needs a new tool.

| Job | What carries over | Where it lives in WordPress |
| --- | ----------------- | --------------------------- |
| Crawlable, indexable pages | Pages must be reachable and not blocked, whoever is asking | Settings, Reading; robots.txt; host and CDN rules |
| Titles and headings | Clear titles, and headings that name the question a section answers | SEO plugin title fields; block editor headings |
| Content that answers the question | Useful, accurate, original; the direct answer comes early | The editor, which is a writing job |
| Author and trust signals | A named author and a real date give any reader, human or model, something to attribute | Author box; Person schema; modified-date settings |
| Page speed and a light page | A lean page costs less for every crawler to read | Caching, theme, plugin count |
| Structured data | Accurate Article and Person markup that matches the visible page; Google says no special AI markup is needed | Your SEO plugin's schema settings |
| Internal links | Links help every crawler find the pages that matter | Block editor; related-posts blocks |

One row deserves a second look. A site that discourages search engines, or whose host blocks crawlers before WordPress answers, is invisible to SEO and GEO alike. It is the cheapest failure to check and the most expensive to miss, so it comes before anything clever.

## What changes, one: more crawlers, each with its own job

Google feeds its AI features from the same index as classic search. The other assistants run separate crawlers for separate purposes, and the name tells you which job a rule controls.

One page, two sets of crawlers

One page, two sets of crawlers

Your
WordPress page
one URL
SEO lane: Google

Googlebot
one shared index
also feeds the AI features

Blue links,
AI Overviews,
AI Mode
GEO lane: everyone else

OAI-SearchBot
PerplexityBot
Claude-SearchBot

ChatGPT Search,
Perplexity,
Claude search

Google serves its AI features from the same index as classic search. The other assistants each run their own crawlers, and robots.txt is where you tell them apart.

| Crawler | Owner | What a rule for it controls |
| ------- | ----- | --------------------------- |
| OAI-SearchBot | OpenAI | Whether ChatGPT Search can find and cite your pages |
| GPTBot | OpenAI | Collection of content for model training, not ChatGPT Search |
| ChatGPT-User | OpenAI | Pages fetched on a user's behalf; OpenAI says robots.txt rules may not apply |
| Claude-SearchBot | Anthropic | Whether Claude's search features can cite you |
| ClaudeBot | Anthropic | Collection of content for model training |
| PerplexityBot | Perplexity | Whether Perplexity's search results can include you |
| Google-Extended | Google | Use for Gemini training and grounding; does not affect inclusion in Google Search |

Two pairs get mixed up constantly. Blocking GPTBot does not remove you from ChatGPT Search, and allowing it does not put you in. Blocking Google-Extended does not remove you from Google Search or its AI Overviews. The vendors describe each crawler in their own documentation, including [OpenAI](https://developers.openai.com/api/docs/bots) and [Anthropic](https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler).

RankReady's list of 29 crawlers covers OAI-SearchBot, GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot and Google-Extended, but not Claude-SearchBot today, so add that one by hand if you want it named.

On WordPress this is a robots.txt job, with two traps. A named group replaces your global rules for that crawler instead of adding to them, which we walk through in [auditing your robots.txt for the most specific match trap](/audit-robots-txt-by-hand/). And your host or CDN may refuse these crawlers before robots.txt is ever read, as covered in [what Cloudflare's crawler default change means for a WordPress blog](/cloudflare-crawler-default-wordpress/). The decision about which crawlers to allow is argued both ways in [blocking or allowing 29 AI crawlers](/block-ai-crawlers-robots-txt/), and [five curl commands](/ai-crawler-test/) will tell you what actually reaches your site.

## What changes, two: the passage is the unit

A search engine ranks a page. An assistant quotes a few sentences from it, so the sentence it picks has to make sense on its own. In the citation analysis behind our other posts, nothing above 18 words was cited, as set out in [nothing longer than 18 words was ever cited](/18-word-ceiling/). That is an observational pattern and not a rule, and it applies to other assistants: Google says you do not need to write in a specific way for its own AI features. If you want to act on it anyway, lead each section with a plain answer sentence and let definitions stand alone.

Ranking and being quoted also come apart, as covered in [ranking first and being the answer are now two different jobs](/ranking-vs-being-the-answer/) and, for ChatGPT specifically, in [what actually changes and what does not](/chatgpt-seo-what-changes/).

## What changes, three: you measure different things

SEO is measured in positions, impressions and clicks. For Google's AI features, Search Console now has the Generative AI performance report. For everything else, the evidence is on your own server: which AI bots fetched which pages, what they received, and which visitors arrived from chatgpt.com, perplexity.ai or claude.ai.

You can read that from an access log by hand, as in [reading your server log for AI crawlers](/read-server-log-ai-crawlers/), or ask an assistant about your own brand and see what comes back, as in [our brand test](/fifteen-second-brand-test/). One caution applies to every bot log. A visit proves the page was fetched, not that it was quoted, so treat it as a leading indicator.

## What changes, four: some advice does nothing yet

For Google, `llms.txt`, AI text files, chunking and manufactured mentions all sit in the ignore pile, and Google says so. For the other assistants the evidence is thin. That is why we wrote that [llms.txt does not work, and we generate it anyway](/llms-txt-does-not-work/): it is cheap, it is harmless, and nobody should expect a result from it.

So the honest summary is short. GEO is not a second discipline. It is SEO, plus access for more crawlers, plus writing that survives being quoted, plus a different way of checking whether it worked.

## The order we would do it in on a WordPress site

First, make sure ordinary SEO is sound, because it feeds Google's AI features directly and it is the part nobody disputes. Second, check that the AI crawlers can reach you at all. Third, give each crawler you care about its own robots.txt group, with your global rules repeated inside it. Fourth, rewrite the answers on your ten most important pages so the plain answer comes first. Fifth, set up a way to see bot visits and AI referrals, so you can tell whether any of it moved.

That list is a summary. The full version, with the evidence behind each step, is in [four jobs for AI search, translated into WordPress](/ai-search-optimization-wordpress/). If you are wondering which name to use for all of this, [GEO, AEO and AI SEO](/geo-aeo-ai-seo/) describe mostly the same work, and the post on that argues why the label matters less than the task list.

## Where RankReady fits

RankReady is a free WordPress plugin for the access and measurement half of this list. It writes robots.txt rules for 29 named AI crawlers, repeating your global rules inside them, serves Markdown copies of your posts, adds Article and Person schema without duplicating what your SEO plugin already writes, and logs which AI bots fetch which pages. It runs alongside Yoast, Rank Math, AIOSEO and SEOPress, which keep owning titles, meta and sitemaps, so there is nothing to migrate. The Markdown copies are for other assistants. Google says it does not use them, so do not expect a change in Google Search.

It does not write your answers for you, and it cannot promise a citation. No tool can. If you want to see where you stand before installing anything, the free [AI Readiness Checker](https://hostmy.blog/plugins/rankready/ai-readiness-checker/) runs 41 checks against your live site, in about fifteen seconds when we last timed it, and needs no account to see the score. The detailed fixes are emailed to you. The plugin itself is on [WordPress.org](https://wordpress.org/plugins/rankready-ai-llm-seo/).

Pick one page that already ranks well. Ask an assistant the question that page answers. Does it name you, and if it does not, which of the four changes above explains why?