What is structured data, and why does your competitor's price appear in Google?
Why is it needed?
The search engine reads the text too, but interpreting text is ambiguous. "$49.90" — is this a price? A discount? A shipping fee? A leftover from an old promotion?
Structured data removes this uncertainty. In a separate block, in a standardized form (with the schema.org vocabulary, in JSON-LD format) you declare: this page is a product, its price is $49.90, it is in stock, and this is the manufacturer. No guessing.
What can you state with it?
| Type | What it states | What it can lead to in the result |
|---|---|---|
| Product (Product) | Name, price, currency, stock, rating | Price and stock under the result |
| Company (Organization / LocalBusiness) | Name, address, contact details, opening hours | Company panel, map display |
| Frequently asked questions (FAQPage) | Question–answer pairs | Expandable questions under the result |
| Article (Article) | Title, author, publication and update date | Date and author in the result |
| Breadcrumb (BreadcrumbList) | The page's position in the structure | A path under the title instead of the raw URL |
| Service (Service) | What you offer, where, at what price | More precise understanding, a fact quotable in AI answers |
What does it give, and what does it not?
What it gives: the appearance of the result. A price, a rating or an expandable question list in the result brings more clicks from the same position. In addition, language models process this layer most reliably — one of the cheapest conditions for getting into AI answers.
What it does not give: ranking. Structured data on its own is not a ranking factor. If the content is bad, structured data won't make it good. And if the data doesn't match what is visible on the page — for example, you declare a different price than what the visitor sees — that is expressly punishable.
The most common mistakes
- None at all. The most common case. In that case the search engine guesses.
- It doesn't match what is visible. A declared price that is not the price shown on the page. This is the most serious mistake.
- Outdated data. Stock says "in stock" although there hasn't been any for weeks. Structured data works when it is generated from the same source as the visible content — not maintained by hand.
- Incomplete required fields. A price without a currency, an article without a date. In that case the rich result simply won't appear.
- The same question on multiple pages. The FAQ block should be page-specific, not identical on every page.
How do you check it?
- Rich Results Test — paste in your own URL, and it shows what Google sees and where it goes wrong.
- Schema Markup Validator — checking against the standard.
- Google Search Console → the Enhancements reports: this shows how many errors there are across the whole site, not just on the one page you happen to be checking.
The goal is not to "have structured data" but to have zero errors. This figure is what is worth requesting in a handover protocol — for me this is basic equipment, example: bucipek.hu case study.
Who builds it?
It works best if it is not something bolted on afterwards but part of the build, generated from the same source as the page's visible content. That way it cannot drift: if the product's price changes, the declared price changes too, because it is the same data.
If your website is being built right now, make this not a separate line item in the quote but a baseline requirement — for me it is: custom website development.
Frequently asked questions
Do you need a plugin for it?
With template-based systems, typically yes, and there the quality depends on the plugin. With custom development, structured data is part of the system, not a separate layer.
Why doesn't it appear in the result even though it's flawless?
Because a rich result is a possibility, not an entitlement: Google decides when to display it. Flawless data is a condition for that, not a guarantee.
Is this the same as the meta description?
No. The meta description is a sentence for humans. Structured data is a machine-facing dataset that the visitor doesn't see.