AI SEO · Schema & Structured Data
Schema markup that states your facts
so nothing is inferred.
Search engines and AI crawlers work out what your pages mean by reading the words around them, and they get it wrong often enough to matter. Structured data hands them the facts directly, in a vocabulary they already parse.
<script type="application/ld+json">
- 1 @type: Organization
- 2 name: LineUp
- 3 url: lineup.agency
- 4 sameAs: verified profiles
The schema.org vocabularies we implement and keep valid
- Organization
- Service
- Product
- Article
- FAQPage
- BreadcrumbList
- LocalBusiness
- Review
Our schema strategy
-
01
Mark up what is true
Declare facts you can evidence, never the ones you would prefer were true.
-
02
Deploy what still earns
Types that produce a result today, and honesty about the ones that no longer do.
-
03
Connect the graph
One identifier per entity, so nodes reference each other instead of competing.
-
04
Validate continuously
Catch breakage the week a template changes, not months after it shipped.
Structured data is a standardised vocabulary, usually written as JSON-LD inside a script tag, that states the facts about a page in a fixed format so search engines and AI crawlers can read them directly instead of inferring them from the surrounding prose.
Your page says it. Your markup proves you meant it.
A crawler reading your about page has to decide whether a sentence names your company, a client or a competitor. Markup removes that decision. It is the cheapest accuracy work available on a website.
Prose leaves room for doubt
Language is ambiguous to a parser. A date could be a publication date or an event date, a number could be a price or a case reference. Markup names each one so nothing is left to interpretation.
One set of facts, two audiences
The visible page is written for a person. The JSON-LD block states the same facts for a machine. When those two versions agree and stay in sync, the crawler has no reason to guess.
It changes understanding, not position
Schema is not a direct ranking factor and we will never sell it as one. What it changes is how your content is understood and how it can be displayed, which is a different lever with a different payoff.
What each schema type actually earns today
Google retired several rich results in 2023 and tightened the rules on others. Plenty of proposals still promise features that no longer exist. Here is the current position on the types most sites deploy, including the ones that pay you back in machine readability rather than in a SERP badge.
| Schema type | What it earns today | Worth deploying? |
|---|---|---|
| Organization | No rich result of its own. It is the record that tells Google and LLM crawlers who publishes the site, what it is called and which profiles belong to it. | Yes. It is the first thing we deploy and the anchor every other node refers back to. |
| Service | No search feature. It states what you sell, where you sell it and what each engagement includes, in a form that can be read rather than summarised. | Yes, for accuracy in AI answers rather than for anything Google will show. |
| Product | Genuine product results in Search and Shopping surfaces, carrying price, availability and shipping detail when the data is accurate and matches the visible page. | Yes, if you sell things. This is the type with the clearest commercial return. |
| Article | No badge on a standard result. It clarifies headline, author, publisher and dates, and supports news eligibility where the other publisher requirements are already met. | Yes, mainly to make authorship and freshness unambiguous. |
| FAQPage | Nothing in the SERP. Google restricted FAQ rich results to government and health authority sites in August 2023, so a commercial site earns no feature from it. | Yes, but only as a machine-readable mirror of your answers for LLM crawlers. Expect no search feature. |
| BreadcrumbList | The breadcrumb trail shown above the page title in place of a raw URL, which also tells crawlers where the page sits in your hierarchy. | Yes. Low effort, still displayed, and useful for site structure on its own. |
| Review | Star ratings only where eligibility is met. Reviews you write or collect about your own business are self-serving and excluded, so most agency and services markup earns nothing. | Only where genuine, qualifying reviews exist. We will tell you when yours do not qualify. |
Who this is for
Structured data pays off wherever a machine has to summarise you without asking you first, and wherever a catalogue is too large to be understood page by page.
Sites that machines describe wrongly
An assistant names your old address, attributes your work to a similarly named firm, or lists a service you dropped years ago. Markup is where the correction starts, because it states the current facts in the one place the crawler cannot misread. Making those facts stick everywhere else is Entity Structuring.
E-commerce catalogues
Product markup is the type with a real, visible payoff, and it is also the easiest to get subtly wrong at scale. Prices that lag the page, stock states that never update and variants that share one identifier all cause the same outcome: the result stops showing.
Publishers and content teams
If you publish under named authors, Article and author markup is how you make the byline machine-readable rather than decorative. It supports the wider work of proving who wrote a piece, when it was last reviewed and what the author is qualified to say.
Brands chasing AI citations
Answer engines quote sources they can parse and verify. Markup is the parse layer underneath Answer Engine Optimization, and it sits on the same technical foundation as your organic SEO. It will not make a weak page worth citing, but it removes the ambiguity that stops a good one from being quoted accurately.
How we implement structured data
Markup is an engineering job with a maintenance cost. We treat it that way, so it keeps validating after the launch week.
-
1
Audit what is already there
We inventory every block currently emitted, run it through validation and record what is broken, duplicated or contradicting the visible page. Most sites have markup from three different plugins fighting each other.
-
2
Model the entity graph
One Organization node, one WebSite node, stable identifiers, and every page-level node referring back to them. Without that spine you get a pile of disconnected snippets instead of a coherent description of a business.
-
3
Map fields to real data
Every property gets bound to a field the CMS already holds, so the markup updates when someone edits a price or a publication date. Hard-coded values are the reason most implementations rot within a year.
-
4
Deploy by template
Markup ships as templates attached to page types rather than as blocks pasted into individual pages. New products and articles inherit correct structured data on publication without anyone remembering to add it.
-
5
Validate and monitor
Every template is checked against the schema.org vocabulary and Google’s own eligibility requirements, then watched in Search Console so an error introduced by a release is caught in days rather than quarters.
-
6
Revisit as the rules change
Supported types get retired and requirements get tightened, as FAQ and HowTo showed. We review what your markup is still earning and retire or repurpose anything that has stopped paying for itself.
What you actually receive
Working markup in your codebase and the documentation to maintain it, handed over in a form your developers can own without us.
The audit and its baseline
- Every structured data block your site currently emits
- Validation errors and warnings, ranked by what they cost you
- Duplicate and conflicting nodes identified by source
- A plain statement of what each type can earn for you
The entity graph
- Organization and WebSite nodes with stable identifiers
- One canonical description reused everywhere it appears
- sameAs references to profiles we can verify
- Page-level nodes linked back to the graph
Deployed templates
- JSON-LD per page type, bound to CMS fields
- Service, Product, Article and BreadcrumbList as they apply
- FAQPage shipped as a machine-readable mirror
- Notes on which properties are required and why
Monitoring and handover
- Search Console enhancement reports configured and explained
- A validation step you can run before each release
- A short maintenance guide written for your developers
- A review when Google changes what a type supports
“We started with a free strategy that showed us specific gaps in our campaigns. Today, LineUp is our regular partner, delivering a measurable return on every invested zloty.”
What clients say on Google
Unedited screenshots from our Google Business Profile. Swipe, or use the arrows, to read them all.
Read these reviews on Google. Reviews written in Polish or Turkish are shown in Google’s own translation.
Questions we get asked
What is schema markup?
Schema markup is code, usually JSON-LD, that labels the facts on a page using the shared schema.org vocabulary. It tells a crawler that a string is a price, an author or a service area rather than leaving it to be inferred from surrounding text.
Will schema markup improve our rankings?
No. Structured data is not a direct ranking factor. It affects how your content is understood and how it can be displayed, which changes click-through and how accurately AI systems describe you, but the markup itself does not lift a position.
Do FAQ and HowTo markup still earn rich results?
Not for a commercial site. Google restricted FAQ rich results to government and health authority sites in August 2023, and removed HowTo rich results the same year. We still ship FAQPage as a machine-readable mirror for LLM crawlers, with no promise of a search feature.
Can you get us star ratings in the search results?
No. Review and rating rich results carry strict eligibility rules, and reviews you collect and publish about your own business are self-serving and not eligible. Where genuine product reviews qualify we mark them up correctly, and we say so upfront when they do not.
How is this different from entity structuring?
Schema is the format used to declare facts on your own pages. Entity structuring is the wider job of making those facts consistent and corroborated everywhere else, across directories, profiles and third-party sources. Schema states the claim, entity work makes it credible.
Do we need a developer to keep it working?
Not usually. Most sites can generate markup from fields already held in the CMS, so it updates when the content does. Hand-pasted blocks drift out of date, which is why we prefer templates plus a validation check that runs on every release.
Find out what your markup is really earning.
We will validate every structured data block on your site, tell you which types still earn something and which are decorative, and send you the fixes ranked by what they are worth.
Last reviewed 28 July 2026.