01Notes · 19 August 2026
What Schema Is Actually For, now that it is not a ranking lever
Quick answer
We tell buyers not to pay for AI-specific schema packages, so we owe them the other half of that sentence: what structured data is genuinely for.
Short version: it does not lift you in rankings or citations. It removes ambiguity about what your page is, which matters, and is a smaller claim than the one usually made for it.
Key takeaways
- No schema type showed a consistent independent citation advantage across 1,645 observations in 2026 research.
- There is no AI-specific schema. Google states this directly. Packages selling it are selling nothing.
- FAQ rich results were deprecated in 2026. FAQ content still helps readers; the display feature is gone.
- What it does do is resolutionstating machine-readably what your visible page already says.
- Five types cover most sites: Organization, Article, BreadcrumbList, Product, and LocalBusiness only if the location is real.
- It can hurt you if it describes something the page does not show. That is a policy problem, not a missed optimisation.
The claims it cannot support
It does not lift rankings by itself, and it does not buy citations. In 2026 research covering 1,645 citation observations, no schema type showed a consistent independent advantage. Pages with rich markup were not reliably cited more than pages without.
There is also no AI-specific structured data. Google says this plainly, and it is worth knowing because a whole category of package is sold on the opposite claim.
And FAQPage markup no longer produces rich results in Google, which were deprecated in 2026. We still use FAQ blocks across this site, because they help a reader find a specific answer and they give an assistant something clean to lift. But building a sitewide strategy around a display feature that no longer displays is effort spent on a memory.
What it does do
It resolves ambiguity. That sounds modest and it is the honest description.
Your page says you are a company called Searlixa based in Jaipur. A human reads that instantly. A machine has to infer it, and inference goes wrong — particularly with similar names, similar companies, or facts that differ across sources. Schema states the same thing in a form that does not need inferring.
That matters most where confusion is likely: a brand name that is also a common word, a person who shares a name with someone better known, a company with several similar entities around it. In those cases removing ambiguity is worth real effort. On a site with an unusual name and consistent facts, the gain is smaller and honest to say so.
| Type | Worth it? | Why |
|---|---|---|
| Organization | Yes | Anchors who you are, once, sitewide. The highest-value item |
| Article / BlogPosting | Yes | Author, dates, publisher. Cheap and useful |
| BreadcrumbList | Yes | States where a page sits. Still displays in results |
| Product | If you sell | Price and availability, where they are genuinely published |
| LocalBusiness | Only if real | A physical location you actually have. Never otherwise |
| FAQPage | Marginal | Rich results deprecated 2026. The content still helps; the markup barely |
| "AI-optimised schema" | Does not exist | Google states this directly |
The one way it can hurt you
Schema must describe what the page visibly shows. Marking up reviews you do not display, a price you do not publish, or an availability you cannot honour is a structured-data policy problem, and the penalty is losing the markup benefits you did earn honestly.
This happens more often by accident than by intent: a template inherits markup for a field that page does not populate, and nobody checks. It is worth a periodic look rather than a one-time implementation.
Every page on this site carries structured data describing exactly what is on it, which is why we can publish a plain fact page for assistants and expect it to be read correctly. That is the whole ambition: not an advantage, just no ambiguity.
What this means for you
- Implement five types and stop. Organization, Article, BreadcrumbList, and Product or LocalBusiness where they genuinely apply.
- Refuse any AI-specific schema package. There is nothing to buy.
- Check your markup matches your visible page, especially on templates nobody has reviewed in a year.
02Common questions
Questions this raises
Does schema help with AI citations?
Not as a direct lever. In 2026 research covering 1,645 citations, no schema type showed a consistent independent advantage. What it does is help systems resolve what your page is about.
Then why implement it at all?
Because it removes ambiguity. Schema states in machine-readable form what your visible page already says, who published it, what it is, when it changed. That helps resolution, which is not the same as ranking.
Is there AI-specific schema?
No. Google says so directly. Any package selling AI-specific structured data is selling something that does not exist.
What about FAQ schema?
Google deprecated FAQ rich results in 2026. FAQ content still helps readers and page clarity, but building a sitewide strategy around a display feature that no longer displays is wasted effort.
Which types are worth the effort?
Organization, Article or BlogPosting, BreadcrumbList, Product where you sell, and LocalBusiness only if you have a real physical location. That covers most sites.
Can schema hurt us?
Yes, in one specific way: if it describes something the page does not actually show. Marking up reviews you do not display, or a price you do not publish, is a policy problem rather than an optimisation.
Sources
Every figure on this page comes from one of the following. Where a claim is ours rather than a source’s, it carries an evidence label instead.
- Rankability — State of AI Search, 1,645 citation observations across 31 topics, 2026. Source for the finding that no schema type showed a consistent independent citation advantage, and for the first-100-words and query-breadth findings. Independent study
- Google Search Central — Optimizing your website for generative AI features (15 May 2026, updated 10 July 2026), structured data documentation, and the 2026 deprecation of FAQ rich results. Primary
- Searlixa sampling across client engagements, labelled Observed with sample and date. Our own
Ask an assistant
Hand this page to an assistant and ask it to summarise or challenge the argument. If it gets something wrong about us, that is itself a useful finding.
→Next step