Rich snippets are enhanced search results showing additional information — star ratings, prices, FAQs, recipe details, event dates. They make your result more visible and generate significantly more clicks than standard blue links.
| Rich result type | CTR impact | Difficulty to implement |
|---|---|---|
| FAQ dropdowns | 🔴 Very high — doubles visual space | Low — add FAQPage schema |
| Star ratings (reviews) | 🟠 High — increases trust signals | Low — add AggregateRating schema |
| Product price/availability | 🟠 High — pre-qualifies buyers | Medium — requires complete Product schema |
| Recipe rich results | 🟠 High — visual card format | Medium — multiple required properties |
| HowTo steps | 🟡 Medium — shows numbered steps | Medium — step-by-step content required |
FAQPage schema is the fastest high-impact rich result for most sites. Add 3-5 questions and answers to any informational page, wrap them in FAQPage schema, validate in the Schema Debugger, and request indexing. FAQPage rich results can appear within days of Google recrawling the page.
The value of a rich result is usually explained as "more clicks", which is true but hides the mechanism — and the mechanism is what tells you which ones are worth pursuing.
A search results page is a competition for vertical space. A plain blue link occupies a fixed, small amount of it. A result with FAQ dropdowns can occupy several times that, pushing the competitor beneath you further down the screen, in some cases off the first viewport entirely. You have not out-ranked them. You have out-sized them, and on a phone that is frequently the more decisive advantage.
The second effect is qualification. Star ratings and prices do not merely attract the eye; they filter it. A result showing a price attracts fewer clicks in total and more clicks from people willing to pay it. For a commercial page that is an improvement even when the raw click-through rate falls, and it is the reason a bare CTR figure is a poor way to judge whether a rich result is working.
This is the single most common misunderstanding about rich results, and it is the reason so many people conclude their schema "does not work".
Valid markup makes a page eligible for a rich result. It does not entitle it to one. Google decides, per query and per page, whether to display the enhancement, and it frequently declines — for pages it does not trust, for queries where the format is unhelpful, or simply because it is testing something. A page can validate perfectly and never show a single star.
Structured data must describe content that is actually visible on the page. Not content that used to be there, not content hidden behind a display:none, not content that exists only in the markup because it would produce a better search result.
This is not a stylistic preference. It is the criterion Google enforces, and enforcement takes the form of a manual action against structured data — at which point every rich result across the entire site disappears, not merely on the offending page. Recovery requires fixing the markup and submitting a reconsideration request, and it takes weeks.
FAQPage is the highest-return rich result and also the one most frequently implemented in a way that eventually triggers a manual action. The failure is nearly always the same: the markup and the page drift apart.
The correct order is to write the FAQ section as content, publish it, and then mark up what is there. The wrong order — and the common one — is to decide which questions would produce a good search result, put them in the JSON-LD, and add a subset of them to the page. That produces a mismatch on day one and it does not self-correct.
"Pricing" is not a question. "How much does a site audit cost?" is. The name property should read exactly as a user would type or speak it, because that is what it is being matched against.
The answer may be extracted with no page around it — read aloud by an assistant, or lifted whole into an AI summary. An answer that begins with throat-clearing is discarded. State the answer, then elaborate.
The most insidious version of the mismatch: someone edits the visible FAQ section a year later, removes a question, and never touches the JSON-LD. The page now claims something it does not show, and nobody knows.
Any informational page can carry it, the content required is content you should have written anyway, and the visual effect is the largest of any common type. The questions must be ones your customers genuinely ask, answered directly in the first sentence — because the answer may be lifted whole into an AI summary with no page around it for context.
Stars are the most trusted visual element on a results page and they carry the strictest rules. The ratings must be real, they must be displayed on the page, and they must not be reviews of yourself.
Price and availability shown in the result pre-qualify the click. The difficulty is that the markup must stay accurate: a price in the search result that does not match the price on the page is both a violation and a rapid way to lose a customer's trust. The markup must be generated from the same data as the page, never hand-written.
Both require genuine step-by-step content with real, complete properties. Neither is a format you should retrofit onto a page that is not actually a set of instructions.
That last point is the strongest argument for not building a strategy on any single rich result type. The formats are Google's to withdraw, and they have withdrawn them before. Markup that also serves answer engines retains its value when a display format is retired; markup that existed only to produce a dropdown does not.
Checks your markup against the schema.org vocabulary. It tells you whether your JSON-LD is syntactically valid and whether the properties exist. It does not know or care what Google requires. A page can pass this cleanly and be ineligible for every rich result Google offers.
Checks eligibility against Google's specific requirements — which are narrower than schema.org's, and which include required properties schema.org treats as optional. This is the one that matters for search display.
Neither tool tells you whether the result will actually be shown. Both will happily pass markup describing content that is not on the page — they check syntax and completeness, not honesty. The rule that gets sites penalised is the one thing no tester enforces, which is precisely why so many sites violate it without realising.
Rich snippets were designed for a search results page. Their value is now shifting somewhere else entirely.
When ChatGPT, Perplexity or an AI overview answers a question using your page, it is reading the page — not querying an index that has already understood it. Structured data is the only part of your page that states, unambiguously and without inference, what a thing is: this is the price, this is the question, this is the answer, this is the rating and it is out of five.
Everything else has to be interpreted from prose, and interpretation is where models make mistakes — quoting a struck-through price, inventing an opening time, attributing a competitor's review to you. Explicit markup is the difference between being cited correctly and being cited wrongly, and being cited wrongly is worse than not being cited at all.
This changes the calculus on markup Google may never display. FAQPage markup that never produces a dropdown is not wasted if it means an answer engine can quote your answer verbatim and attribute it to you.
Valid markup makes a page eligible; it does not entitle it to a result. Google decides per query and per page whether to display the enhancement, and shows them more readily to sites it already trusts.
No, and this is the most common cause of a structured-data manual action. Markup must describe content a reader can actually see. A manual action removes rich results across the whole site, not just the offending page.
No. Self-serving reviews are explicitly disallowed, regardless of whether the review is genuine.
FAQPage. It applies to almost any informational page, requires content you should have written anyway, and produces the largest visual gain. It is also the safest, provided the questions genuinely appear on the page.
They matter more. AI answer engines read the page itself rather than a pre-built index, and structured data is the only part of a page that states unambiguously what a thing is.
Run the Rich Snippets Tester and get results in minutes. Pay as you go.
Test Rich Snippets →aiwebpageseo.com is a data-driven SEO and AEO (Answer Engine Optimisation) platform providing a free suite of technical website tools. Rather than relying on AI-theorised assumptions, the platform analyses live URL performance, delivering objective diagnostics, page speed metrics, CLS debugging, and site crawl data alongside actionable technical tutorials.