Getting cited by ChatGPT or a Google AI Overview does not come from adding a plugin and checking a box. It comes from three things working together on the same WordPress site: structured data that removes ambiguity about who wrote the page and who published it, content written so a single paragraph can be lifted out and still make sense, and pages that get revisited and updated instead of published once and left alone. Skip any one of the three and the other two do a lot less work.
That is a different problem than classic SEO, and it is worth being specific about why, because a lot of what is being sold as “AI SEO” right now is the old checklist with a new label stapled on.
Schema markup alone does not get you cited
This is the part most guides skip, because it is not a flattering finding for the plugins selling schema as a silver bullet. Ahrefs tracked 1,885 pages that added JSON-LD structured data between August 2025 and March 2026 and found no statistically significant citation uplift from adding it in isolation, across Google AI Overviews, AI Mode, or ChatGPT.
That does not mean schema is a waste of a WordPress dev’s time. It means schema is infrastructure, not a growth lever on its own. Organization, Article, and Person markup tell a language model unambiguously who ConicPlex is, who wrote a given page, and how the page relates to the rest of the site. That removes friction for a model deciding whether to trust a source. It does not manufacture trust that the content itself has not earned. If the underlying page is thin or the byline is fake, adding a JSON-LD block on top will not fix that.
Where schema earns its keep is downstream of everything else: after the content is genuinely worth citing, after the authorship is real, structured data is what makes that clear to a machine reading the page in milliseconds instead of a human reading it for two minutes.
What that looked like on a real build
We built EEAT WP Pro for a client, Derek Gaugan, who had the concept but needed it turned into an actual shipped WordPress plugin: a free and premium tier, author biography display with schema-rich markup, trust signal badges, credential links, and a subscription licensing system with its own admin dashboard for tracking who is on which plan. Four weeks from brief to a plugin live and deployed on WordPress.org, the kind of build our plugin development team handles when a client has the concept and needs it turned into something people can actually install. There’s a broader library of similar work in the plugins portfolio if the shape of that kind of build is useful context.
The reason that build is relevant here isn’t that a plugin now exists. It’s that shipping it required deciding, concretely, which E-E-A-T signals are worth exposing as structured data (author identity, credentials, publisher identity) and which ones are really a content and editorial problem no plugin can automate. That split matters more than most “install this and rank” pitches let on, and it’s the same split that shapes the checklist further down this page.
How is AI citation different from ranking in search results?
Ranking in Google’s ten blue links rewards a page that satisfies a query well enough to earn a click. Citation in an AI answer rewards a passage that can be lifted out of a page, dropped into a generated answer, and still be correct and complete on its own. A page can rank fine and get cited zero times if every useful sentence depends on the three sentences before it for context.
Practically, that means writing fact-dense, self-contained passages in the body of a post, not just a strong opening. It’s the same extraction-first thinking that shows up in AI development work generally: a model can only act on what it can parse cleanly, whether that’s a paragraph feeding an answer engine or a document feeding a retrieval pipeline. A paragraph buried in section four should still make a real, specific claim, not a transition sentence that only means something next to its neighbors. If a sentence needs the whole page to make sense, an extraction engine will usually skip it in favor of a competitor’s page where the same idea stands on its own.
Freshness matters more for ChatGPT than almost anything else
Ahrefs’ analysis of ChatGPT’s most-cited pages found that AI-cited content runs about 25.7% fresher on average than pages that simply rank well organically, and the gap is sharper for ChatGPT specifically: 76.4% of its most-cited pages had been updated within the previous 30 days. Recently updated pages in that dataset averaged 6 citations against 3.6 for outdated ones, a 67% gap that has nothing to do with the content being better written and everything to do with it being current.
For a WordPress blog with a backlog of evergreen posts, that changes the priority order. Writing one more new post this month is worth less, for citation purposes, than going back into three older cornerstone posts, checking whether the facts and numbers in them still hold, and updating the ones that don’t. A pricing post from eight months ago with stale numbers is actively working against a site’s citation rate, not just sitting there neutrally.
A working checklist, in the order it actually pays off
- Organization and Person schema first. This is the highest-leverage, lowest-effort step: it tells language models unambiguously who is publishing and who wrote each page.
- Real author bios on every post, not a generic “Admin” byline, with credentials that are actually true. This is the E-E-A-T signal that plugins can format but cannot fabricate.
- Rewrite the first 60 to 100 words of key pages as a direct, self-contained answer to the question the page targets, not a warm-up paragraph before the real content starts.
- Audit your ten highest-traffic older posts for stale numbers, dead links, or outdated advice, and update them with a visible revision date before writing anything new.
- Add FAQPage schema only where there’s a real FAQ, not a bolted-on section written to hit a checklist item.
None of that is exotic. What’s different from 2023-era SEO advice is the order: authorship and freshness now do more of the work that keyword density and backlink counts used to do, and schema is the layer that documents the work rather than the thing that creates it.
Frequently asked questions
Does adding schema markup guarantee my WordPress site gets cited by ChatGPT?
No. Ahrefs’ tracking of pages that added JSON-LD found no statistically significant citation increase from schema alone. Schema clarifies authorship and publisher identity for machines; it doesn’t manufacture trust in content that hasn’t earned it through real authorship and freshness.
What’s the fastest E-E-A-T improvement for an existing WordPress blog?
Real author bios with genuine credentials, paired with Organization and Person schema. It’s a few hours of work per author, not a rebuild, and it directly addresses the authorship signal that both AI answer engines and Google’s own guidelines weight heavily.
Do I need a dedicated SEO plugin to add E-E-A-T schema, or can it be done manually?
Either works. Rank Math and similar plugins handle Article and Organization schema out of the box; author-level Person schema and custom trust signals usually need either a plugin built for that specific purpose or a developer adding structured data directly to the theme’s author templates.
How often should older posts be updated for AI citation freshness?
There’s no fixed interval, but the data points toward quarterly review of a site’s top-performing evergreen content at minimum. ChatGPT’s most-cited pages skew toward updates in the last 30 days, so a post that hasn’t been touched in six months or more is at a real disadvantage regardless of how good the original writing was.
