Do we need any paid plugins to run the scan?

No — and you don’t need one to fix it either. The scan runs externally on GEO Intel: it installs nothing on your site, needs no login, and reads the HTML your server already sends to any visitor or crawler. The remediation is the same story — we ship schema and markup in a child theme and functions.php, so nothing we build depends on a Yoast Premium, Rank Math Pro or Schema Pro subscription. No plugin required to audit, no paid plugin required to fix.

Does WordPress block AI crawlers?

Not by default. WordPress serves your posts and pages as rendered HTML that GPTBot, ClaudeBot and PerplexityBot can read on the first request. The blocks are usually bolted on: a security plugin like Wordfence, a “Discourage search engines” setting left on, a host- or Cloudflare-level bot rule, or a locked-down robots.txt a previous developer left behind. The scan checks every one of those and tells you which are shutting the engines out.

Do you edit my theme — will a WordPress update wipe your work?

No, because we don’t touch your parent theme or core. Every change ships in a child theme — its own functions.php, template overrides and Gutenberg blocks — which loads on top of your active theme and survives its updates intact. We build and preview on staging first, and you approve the go-live. Update your theme next month and the AI-visibility work stays exactly where it is.

Will this hurt my Google SEO?

No — it reinforces it. Valid server-rendered Article, FAQPage and Breadcrumb schema drives Google rich results, and answer-first copy improves both AI citation and classic rankings. We don’t change your URL structure, strip canonical tags, or remove your existing SEO plugin’s output where it’s correct. The foundational signals overlap, so the AI work and your SEO pull in the same direction.

I use Elementor / Divi — is that a problem?

It’s a common one, not a blocker. Page builders wrap your content in deeply nested divs, inline styles and JavaScript widgets that can bury the passage an engine wants to quote, and some modules render content in ways crawlers read poorly. We don’t make you rebuild the site — we restructure the templates and inject clean, server-rendered schema so the answer is machine-readable while the page looks identical to visitors. WPBakery, Beaver Builder and the block editor are handled the same way.

Do I have to uninstall Yoast or Rank Math?

No. Keep them — they handle titles, sitemaps and redirects perfectly well. Where they tend to go light is the schema depth AI engines weight: an authored Person with credentials, a complete FAQPage, sameAs authority links, some of which needs their premium tiers. We add that layer in code and make sure it doesn’t duplicate or collide with your plugin’s output, so you get full-depth schema without a second subscription.

How is this different from setting up WordPress SEO?

WordPress SEO — Yoast or Rank Math plus keywords and meta tags — gets you ranked in Google’s blue links. This gets your pages named inside an AI-generated answer when someone asks a question you answer. The signals differ: AI citation depends on server-rendered schema depth, author identity and passage-extractable copy, not on backlinks or a green readability light. A site can rank page one and still be absent from every ChatGPT answer.

Does this work on headless WordPress?

Yes. On a headless setup — WordPress as the backend with a Next.js, Astro or similar front end over WPGraphQL or the REST API — we ship the JSON-LD and metadata in your front-end routes instead of a child theme, and we confirm the front end server-renders content rather than leaving it to client-side JavaScript the crawlers skip. The audit itself is stack-agnostic: it inspects whatever HTML is publicly served, headless or classic.

A WordPress site audit starts at $799 and covers the full 7-engine scan with a severity-ranked, WordPress-specific remediation playbook. Implementation — where we build the child theme, schema and content fixes — is fixed-price, scoped from the findings to your theme and page count. No retainer, no per-plugin fee, and the 60-day re-scan that measures the citation delta is included in the implementation engagement.

How long does it take?

The audit lands in 48 hours from kickoff. Implementation typically ships within two weeks, depending on your theme, page-builder setup and how many pages need answer-first rewrites. Schema and crawlability fixes register with AI engines within days of going live; content-driven citation movement usually shows within 30 to 60 days, which is what the re-scan captures.

WordPress AI Visibility — Get Cited in ChatGPT & AI Search

WordPress · GEO / AEO / AIO

← All services

When a buyer asks ChatGPT or Perplexity a question you answer better than anyone, the engine reads the HTML your server sends and the schema inside it — not the plugins bolted on top. We scan your site across 7 AI engines with nothing installed, then ship the fixes in a child theme, functions.php and Gutenberg blocks, so your visibility never depends on a paid SEO plugin you keep renewing.

Book a WordPress audit View live scanner

7AI engines tested
0plugins to install for the scan
100+WordPress signals audited
48haudit to scored report

Why AI visibility matters

Two in five sites run WordPress. AI reads almost none of them well.

AI referral traffic is compounding fast, and the engines pick their own citations — mostly sources that never ranked in Google’s top 10. If a crawler can’t cleanly read your page, you are not in the answer it hands the buyer.

~43%Share of all websites that run on WordPress (over 60% of the CMS market) — a figure that has since eased to roughly 41.5%W3Techs, 2025 (via WordPress.com / The Repository)
1.08%AI assistants’ share of all website referral traffic by mid-2025, up from ~0.02% a year earlier — more than a 7x jump, still climbing ~1% month over monthSE Ranking / Analyze.ai AI-traffic studies, 2025
80%Sources cited in AI-search answers that do not appear in Google’s traditional top 10 for the same query — only ~12% match exactly. AI picks its own citationsOmnibound AI Search Statistics 2025–2026
97%Share of llms.txt files that received zero AI-bot requests across 137,000 sites — evidence the major engines ignore the file todayAhrefs study, May 2026 (via PPC.land)
~78%ChatGPT’s share of all AI referral visits; Perplexity drives ~15% globally and near 20% in the US — the platforms a WordPress site is realistically cited inSE Ranking AI-traffic research, 2025
21MB+Unzipped code Elementor alone can add to an install; Gutenberg builds hit median mobile LCP under 2.0s vs 3.8–5.2s for pre-optimized Elementor pagesCoreWebVitals.io WordPress guide, 2026

The WordPress advantage

WordPress starts ahead — then stalls at the schema.

WordPress renders PHP server-side, so your post reaches an AI crawler on the first request — an edge client-rendered stacks don’t get for free. The gap opens later: page-builder markup that can bury the answer, and baseline plugin schema that leaves out the author signals engines weight.

Your siteWordPress posts & pages
Every URL WordPress serves to a visitor — the same HTML a crawler receives.
Server renderPHP → HTML + JSON-LD
WordPress renders PHP into static HTML and the structured data on the server — before any script runs.
No-JS fetchAI crawler
GPTBot, ClaudeBot and PerplexityBot read the raw HTML. They do not execute JavaScript or run your plugins.
CitationCited answer
The engine names your page — author, answer and freshness intact — inside the buyer’s answer.

Where WordPress starts ahead

WordPress renders server-side PHP into static HTML, so an AI crawler that never runs JavaScript reads your post — heading, answer and body copy — on the very first request.
Structured data is not a paid feature: Yoast free and Rank Math free both auto-emit JSON-LD for WebPage, Article, Organization / Person and BreadcrumbList — the foundation is already there at zero cost.
Because WordPress is content- and author-heavy, real author archives plus complete Person schema carry E-E-A-T weight that commerce platforms — where author identity is usually absent — can’t offer AI at all.

Where the default stack stalls

Page builders — Elementor, Divi, WPBakery — wrap content in hundreds of nested <div> wrappers and load CSS/JS for every widget whether used or not, bloating the DOM, hurting Core Web Vitals and degrading passage extraction.
The free SEO plugins ship a baseline Article and Organization block; the authored Person, credentials and sameAs links AI weights for E-E-A-T often need a premium tier or custom code to complete.
WordPress serves a virtual robots.txt by default, so you have no explicit control over which AI bots may read or train on your content until you replace it with real rules.
llms.txt is ignored by every major answer-engine crawler today — 97% of published files get zero AI hits and Google confirms no AI service even checks for it — so WordPress visibility rides on served HTML plus complete JSON-LD, nothing else.

/ 01Why us

What we bring to WordPress AI optimisation.

An audit is only useful once the fixes ship. We built the scanner, we know the theme internals, and we implement in code — server-rendered so the engines that can’t run JavaScript still read your pages, without a plugin you have to keep renewing.

We run the scanner — it’s our own SaaS
The audit runs on GEO Intel, our own GEO platform at geo.atellius.com — the same 7-engine scoring engine we built and operate. You are not paying us to run someone else’s tool and read the output back to you. When a check needs a WordPress-specific rule — a Yoast schema quirk, an Elementor render pattern — we add it to the scanner ourselves.
We fix in code — no paid-plugin lock-in
No plugin required to audit. No paid plugin required to fix. We ship JSON-LD, author schema and answer-first markup through a child theme, functions.php and native Gutenberg blocks — so your AI visibility never sits behind a Yoast Premium, Rank Math Pro or Schema Pro subscription. Stop renewing the licence and the work stays live.
Scored across 7 engines, not one blended number
We test ChatGPT, Perplexity, Google AI Overviews, Gemini, Copilot, Claude and Grok — because a buyer asking the same question gets a different source list on each. A schema fix that wins Perplexity can do nothing on AI Overviews. You get a per-engine score out of 100, not one average that hides where you lose.
Author, E-E-A-T and Article / FAQ schema at full depth
Most SEO plugins emit a baseline Article and Organization block — a solid start. The authored Person, credentials and sameAs links that prove who wrote a post — the E-E-A-T signals AI engines weight — often need a paid tier or custom code to complete. We build the full graph in code: Article tied to a real Person, FAQPage on the questions you answer, BreadcrumbList and Organization, all server-rendered. That is the difference between a passing mention and a cited source.
We tame page-builder bloat
Popular page builders like Elementor and Divi add deeply nested markup, inline styles and JavaScript widgets that can bury the one paragraph an engine wants to quote. We restructure the templates so the answer sits in clean, server-rendered HTML a crawler can lift verbatim — without rebuilding your site. The page looks identical to visitors and reads cleanly to machines.

/ 02How it works

Scan, fix in code, prove the delta.

Five stages from an external site scan to a 60-day re-scan on the same baseline. Every change ships in a child theme and previews on staging; your parent theme and core stay untouched.

01
WordPress site scan
GEO Intel scans your live site from the outside — it installs nothing, needs no login, and reads the HTML WordPress already serves, scoring 100+ signals across all 7 engines. You get an overall score out of 100, per-engine readiness, and a findings inventory ranked by projected citation impact, not by plugin.
Days 1–3
02
Child theme & code remediation
We set up a child theme so every change survives your next theme update, then fix the structural blockers in code — bot access in robots.txt, canonical and heading hygiene, and any content a builder was injecting via JavaScript the crawlers never run. Your parent theme and core stay untouched; you preview on staging before anything goes live.
Days 4–9
03
Schema build in functions.php
We emit the full JSON-LD graph from the child theme’s functions.php on the wp_head hook — Article with an authored Person and credentials, FAQPage, BreadcrumbList and Organization — mapped to your real author and post data so it stays accurate as you publish. Every block validates in Google’s Rich Results Test on the first pass. No Schema Pro licence involved.
Days 6–11
04
Content & Gutenberg blocks
We rewrite posts and key pages answer-first — one concrete claim per passage an engine can quote verbatim — and add reusable Gutenberg blocks for FAQs and answer boxes that carry their schema with them. Marketing filler and unattributed superlatives get cut.
Days 10–16
05
Re-scan & citation delta
At 60 days we re-run the same 100+ checks and the same buyer prompts across all 7 engines and hand you the delta: which findings closed, how per-engine scores moved, and where your pages now get cited that they weren’t before. Same baseline, measured before and after.
Day 60

/ 03The checklist

What your WordPress site should do.

Fifteen checks, ordered by citation impact. We show you what to fix and why it moves the needle — the exact directives for your site, mapped to your theme, come from the scan.

Fix #6 · the crawler switch · functions.php robots_txt filter

Rewrite the virtual robots.txt.

WordPress serves a virtual robots.txt you don’t control. Filter the robots_txt hook from a child theme’s functions.php — or drop a physical /robots.txt, or use Yoast / Rank Math’s editor — to explicitly allow every AI answer-engine token and disallow the scrapers that only cost you bandwidth.

Anthropic alone runs three agents — ClaudeBot for training, plus Claude-SearchBot and Claude-User for the searches that actually surface citations. Allowlist one and you miss the other two.

functions.php

<?php// child theme functions.php — rewrite the virtual robots.txtadd_filter( 'robots_txt', function ( $output, $public ) {   // Allow the AI answer-engine crawlers that cite you  $allow = array(

The complete filter — every answer-engine token in priority order, the physical-file and Yoast / Rank Math variants, and the server-layer scraper blocks Bytespider ignores — ships with your scan report, mapped to your theme.

Scan your site free

01High effortHigh impact
Audit served HTML for page-builder div-soup
AI extractors favor clean, shallow, semantic HTML; Elementor, Divi and WPBakery bury answers in deeply nested div wrappers and render-blocking assets that hurt both extraction and Core Web Vitals.
Exact fix in your scan report
02Low effortHigh impact
Write answer-first content in Gutenberg
AI lifts self-contained passages, so each page should open with a direct 40–60 word answer under a question-phrased heading rather than burying it below marketing copy.
Exact fix in your scan report
03Low effortMedium impact
Complete core JSON-LD with a free SEO plugin
Explicit Article, Organization / Person and BreadcrumbList schema grounds your entities for machines and is fully available at zero cost.
Exact fix in your scan report
04Low effortMedium impact
Add FAQPage / HowTo markup where genuinely applicable
Even after Google dropped the rich-result display, the markup plus a visible Q&A structure still helps AI parse discrete question-answer pairs.
Exact fix in your scan report
05Medium effortMedium impact
Build real author pages with Person schema
WordPress is content- and author-heavy, and AI weighs author experience, expertise and authoritativeness (E-E-A-T) when choosing whom to cite.
Exact fix in your scan report
06Low effortMedium impact
Replace the virtual robots.txt with explicit AI-crawler rules
By default WordPress serves a virtual robots.txt, so you have no explicit control over which AI bots may read or train on your content.
Exact fix in your scan report
07Low effortLow impact
Block rogue scrapers at the server layer
Bytespider and similar scrapers ignore robots.txt Disallow rules, so a robots.txt entry alone will not stop them.
Exact fix in your scan report
08Low effortMedium impact
Install a caching + optimization plugin for Core Web Vitals
Slow LCP / INP wastes crawl budget, weakens UX signals, and correlates with lower AI-answer inclusion.
Exact fix in your scan report
09Low effortMedium impact
Optimize and modernize images
Images are usually the LCP element and a major weight driver on WordPress pages.
Exact fix in your scan report
10Medium effortMedium impact
Prune plugin bloat and dequeue unused assets
Every active plugin can add CSS, JS and DB queries site-wide, slowing pages that don’t even use the feature.
Exact fix in your scan report
11Low effortMedium impact
Generate and submit an XML sitemap
A clean sitemap speeds discovery of new and updated content by both search and AI crawlers.
Exact fix in your scan report
12Medium effortMedium impact
Fix internal linking and clean up taxonomy
Descriptive in-content links help AI map your entities and topical authority, while thin, auto-generated tag archives dilute crawl focus and create duplicate thin pages.
Exact fix in your scan report
13Low effortMedium impact
Set Organization / LocalBusiness schema with NAP
Consistent entity data grounds brand mentions so AI attributes citations to the right organization — and the right location.
Exact fix in your scan report
14High effortHigh impact
Server-render content on any headless build
A client-only React / SPA front end can hide primary copy from AI extractors, negating WordPress’s normally crawlable output.
Exact fix in your scan report
15Low effortLow impact
Enable breadcrumbs and enforce clean heading hierarchy
BreadcrumbList schema plus a strict H1 ▸ H2 ▸ H3 order helps AI understand page context and content structure.
Exact fix in your scan report

The fixes are in the report

These are the checks. Your site’s exact fixes are in the scan.

The audit reads what your server already sends — you get the directives, the priority order and the code, scoped to your theme and page count.No plugin required to audit. No paid plugin required to fix.

Scan your site free

/ 04Before / after

The same query. A different source list.

A buyer asks the assistant a question your best post answers. Before, the engine reads two competitors’ markup and yours is invisible. After, your article carries a full server-rendered graph — Article tied to a named author, FAQPage, freshness — and gets cited first.

PerplexityBefore

› best accounts-payable automation for mid-market

Most roundups point to Bill.com and Tipalti — both publish detailed comparison pages with clear feature breakdowns.

Sourcesbill.comtipalti.com

Your article — not mentioned

PerplexityAfter

› best accounts-payable automation for mid-market

A frequently cited primer is Ledgerwise’s guide to AP automation for mid-market finance teams — an authored, recently updated breakdown of the buying criteria by company size.

AP Automation for Mid-Market Finance TeamsBy Dana Osei, CPAUpdated Jun 20269-min read

Sourcesledgerwise.iobill.com

Your article — cited first

The after answer is only possible because the Article is tied to a real Person with credentials, the FAQPage and answer-first copy are server-rendered, and the /author archive proves who wrote it — none of it injected by a plugin the crawler skips.

/ 05Pricing

Priced per site. No retainer, no paid-plugin lock-in.

Start with the audit to see where you stand, then scope the code remediation from the findings. Everything we ship lives in your child theme — not behind a licence.

WordPress AI Audit

from $799

One 48-hour scan across all 7 engines. Scored report, per-engine readiness, and a severity-ranked, WordPress-specific remediation playbook — no plugin installed, no login needed.

Best for: a first read on where your site stands before committing to code work.

Before and after, on the same prompts.

Each re-scan runs the same buyer queries on the same engines as the initial audit. The delta is measured, not estimated.

Ledgerwise BlogB2B SaaS blog
21/10069/100
Named by ChatGPT in 5 of 8 test prompts for “best accounts-payable automation for mid-market” within 60 days — zero mentions before the audit.
Fix appliedArticle + Person + FAQPage JSON-LD shipped from a child theme functions.php, extending the site’s default schema, plus answer-first rewrites on the top 12 posts.
Marrow & ValeLocal law firm
14/10066/100
Google AI Overviews now cites the firm on 3 “employment lawyer [city]” queries, and Perplexity names it where it previously named a legal directory.
Fix appliedAttorney and FAQPage schema per practice-area page via functions.php, robots.txt opened to GPTBot and PerplexityBot, and Elementor markup flattened to server-rendered HTML.
The Vital BriefHealth publisher
18/10064/100
Perplexity cites the publisher on 6 of 9 tracked condition queries, and Gemini references named-author articles it previously skipped.
Fix appliedPerson schema with author credentials and sameAs, MedicalWebPage on treatment pages, Article dateModified surfaced, and Divi page-builder bloat stripped from the article template.

Site names anonymised; scores and prompts are the client’s own before / after audit runs.

Cited across every engine your buyers ask

ChatGPT
Perplexity
Claude
Gemini
Copilot
Grok

/ 07Common questions

What WordPress owners ask first.

Do we need any paid plugins to run the scan?
No — and you don’t need one to fix it either. The scan runs externally on GEO Intel: it installs nothing on your site, needs no login, and reads the HTML your server already sends to any visitor or crawler. The remediation is the same story — we ship schema and markup in a child theme and functions.php, so nothing we build depends on a Yoast Premium, Rank Math Pro or Schema Pro subscription. No plugin required to audit, no paid plugin required to fix.
Does WordPress block AI crawlers?
Not by default. WordPress serves your posts and pages as rendered HTML that GPTBot, ClaudeBot and PerplexityBot can read on the first request. The blocks are usually bolted on: a security plugin like Wordfence, a “Discourage search engines” setting left on, a host- or Cloudflare-level bot rule, or a locked-down robots.txt a previous developer left behind. The scan checks every one of those and tells you which are shutting the engines out.
Do you edit my theme — will a WordPress update wipe your work?
No, because we don’t touch your parent theme or core. Every change ships in a child theme — its own functions.php, template overrides and Gutenberg blocks — which loads on top of your active theme and survives its updates intact. We build and preview on staging first, and you approve the go-live. Update your theme next month and the AI-visibility work stays exactly where it is.
Will this hurt my Google SEO?
No — it reinforces it. Valid server-rendered Article, FAQPage and Breadcrumb schema drives Google rich results, and answer-first copy improves both AI citation and classic rankings. We don’t change your URL structure, strip canonical tags, or remove your existing SEO plugin’s output where it’s correct. The foundational signals overlap, so the AI work and your SEO pull in the same direction.
I use Elementor / Divi — is that a problem?
It’s a common one, not a blocker. Page builders wrap your content in deeply nested divs, inline styles and JavaScript widgets that can bury the passage an engine wants to quote, and some modules render content in ways crawlers read poorly. We don’t make you rebuild the site — we restructure the templates and inject clean, server-rendered schema so the answer is machine-readable while the page looks identical to visitors. WPBakery, Beaver Builder and the block editor are handled the same way.
Do I have to uninstall Yoast or Rank Math?
No. Keep them — they handle titles, sitemaps and redirects perfectly well. Where they tend to go light is the schema depth AI engines weight: an authored Person with credentials, a complete FAQPage, sameAs authority links, some of which needs their premium tiers. We add that layer in code and make sure it doesn’t duplicate or collide with your plugin’s output, so you get full-depth schema without a second subscription.
How is this different from setting up WordPress SEO?
WordPress SEO — Yoast or Rank Math plus keywords and meta tags — gets you ranked in Google’s blue links. This gets your pages named inside an AI-generated answer when someone asks a question you answer. The signals differ: AI citation depends on server-rendered schema depth, author identity and passage-extractable copy, not on backlinks or a green readability light. A site can rank page one and still be absent from every ChatGPT answer.
Does this work on headless WordPress?
Yes. On a headless setup — WordPress as the backend with a Next.js, Astro or similar front end over WPGraphQL or the REST API — we ship the JSON-LD and metadata in your front-end routes instead of a child theme, and we confirm the front end server-renders content rather than leaving it to client-side JavaScript the crawlers skip. The audit itself is stack-agnostic: it inspects whatever HTML is publicly served, headless or classic.
What does it cost?
A WordPress site audit starts at $799 and covers the full 7-engine scan with a severity-ranked, WordPress-specific remediation playbook. Implementation — where we build the child theme, schema and content fixes — is fixed-price, scoped from the findings to your theme and page count. No retainer, no per-plugin fee, and the 60-day re-scan that measures the citation delta is included in the implementation engagement.
How long does it take?
The audit lands in 48 hours from kickoff. Implementation typically ships within two weeks, depending on your theme, page-builder setup and how many pages need answer-first rewrites. Schema and crawlability fixes register with AI engines within days of going live; content-driven citation movement usually shows within 30 to 60 days, which is what the re-scan captures.

Not on WordPress?

See our full GEO / AEO / AIO audit — for any stack →

Find out what AI says when a buyer asks the question you answer.

Most sites don’t know they’re invisible to AI search until a competitor shows up in their buyer’s ChatGPT answer instead. The 48-hour audit scores your site across 7 engines and hands you the exact child-theme, functions.php and content fixes to ship — no paid plugin required.

Book a WordPress audit Explore the GEO service

AIanswerenginesskipyourWordPresssite.Wefixitincode.

Two in five sites run WordPress. AI reads almost none of them well.

WordPress starts ahead — then stalls at the schema.

What we bring to WordPress AI optimisation.

We run the scanner — it’s our own SaaS

We fix in code — no paid-plugin lock-in

Scored across 7 engines, not one blended number

Author, E-E-A-T and Article / FAQ schema at full depth

We tame page-builder bloat

Scan, fix in code, prove the delta.

WordPress site scan

Child theme & code remediation

Schema build in functions.php

Content & Gutenberg blocks

Re-scan & citation delta

What your WordPress site should do.

Rewrite the virtual robots.txt.

Audit served HTML for page-builder div-soup

Write answer-first content in Gutenberg

Complete core JSON-LD with a free SEO plugin

Add FAQPage / HowTo markup where genuinely applicable

Build real author pages with Person schema

Replace the virtual robots.txt with explicit AI-crawler rules

Block rogue scrapers at the server layer

Install a caching + optimization plugin for Core Web Vitals

Optimize and modernize images

Prune plugin bloat and dequeue unused assets

Generate and submit an XML sitemap

Fix internal linking and clean up taxonomy

Set Organization / LocalBusiness schema with NAP

Server-render content on any headless build