How to Win Featured Snippets and Rich Results

Two very different things get lumped together under "SEO snippets," and mixing them up wastes effort. A featured snippet is the answer box Google builds by lifting a passage from a page that already ranks — you win it with structure, not code. A rich result is a visual upgrade to a normal search listing — star ratings, an FAQ dropdown, a recipe card — and it only appears because the page carries structured data telling Google what kind of content it is. This guide covers how to structure a page to win each type of featured snippet, and where structured data fits in for rich results.

Featured snippets vs rich results — not the same thing

It is worth being precise here, because the fix for one does nothing for the other.

A page can win a featured snippet with zero structured data — Google is reading your prose, headings, lists, and tables, not a schema block. A page can also carry perfect FAQ or Product schema and never appear as a featured snippet, because rich results and snippets are judged independently. They reinforce each other in practice — a page clear enough to be extracted is usually also worth marking up — but they are two separate jobs.

The three featured snippet types, and how to structure for each

Google extracts snippets in three shapes. Which one it pulls depends on the nature of the answer, not on which type you would prefer — so structure the content the way the answer naturally wants to be read, and let Google pick the format.

Paragraph snippets

The most common type. Google lifts a short block of text — typically 40 to 60 words — that directly answers a question. To be extractable, that answer needs to sit immediately after a heading that states the question plainly, with nothing padding it out first.

## What is a redirect chain?

A redirect chain is a sequence of two or more redirects between the
original URL and the final destination page. Each extra hop adds
latency, dilutes link equity slightly, and increases the chance a
crawler gives up before reaching the final page.

Notice what is missing: no "Great question! Let's dive in," no scene-setting paragraph before the definition. The answer comes first; the elaboration, examples, and nuance follow after.

List snippets

When a query implies steps or a set of items — "how to," "types of," "best" — Google prefers to extract a real ordered or unordered HTML list (<ol> or <ul>), not a paragraph that merely reads like a list. Steps use <ol>; unordered collections use <ul>.

## How to fix a broken internal link

1. Crawl the site to find every page linking to the broken URL.
2. Decide whether the destination should redirect or be restored.
3. Update the source pages to point at the correct URL directly.
4. Re-crawl to confirm no page still links to the dead URL.

Writing "First, you'll want to... Next, you should..." as flowing prose might read fine to a human, but it gives Google nothing clean to lift. If the content is genuinely a sequence, mark it up as one.

Table snippets

Comparison data — pricing tiers, spec differences, before/after figures — extracts best from an actual <table> element with clear header cells, not from a bullet list dressed up to look tabular. Keep the table narrow and legible: a handful of columns Google can render cleanly in a small results box, not a wide reference table with a dozen columns.

Answer-first writing: the one habit that wins all three

Every example above shares the same underlying discipline, often called answer-first or inverted-pyramid writing: pose the question as the heading, answer it in the very next sentence, then expand with detail, caveats, and examples afterward. This is the opposite of how much web copy is written — build-up, then the point — and it is why genuinely good content still loses the snippet to a shorter, blunter competitor.

The pattern to follow for any section you want extracted:

1. Heading — the actual question, phrased the way a person would ask it.
2. Immediate answer — one to three sentences, self-contained, no pronouns referring back to something above.
3. Expansion — examples, exceptions, depth, for the reader who wants more than the extract.

This is a narrower, more mechanical version of the intent-matching covered in Search Intent and Content — matching format to intent gets you ranking; answer-first structure within that format is what makes a specific passage liftable once you are.

The same structure now wins AI Overviews and AI answer engines too

This is not just about the classic featured snippet box any more. Google AI Overviews, and AI answer engines like ChatGPT and Perplexity, work the same way structurally: they favour content that is unambiguous, self-contained, and easy to lift out of context — a clear question paired with a direct answer, not a paragraph that only makes sense read alongside three others. Optimising for extraction has effectively become optimising for two audiences at once, and it is increasingly called generative engine optimisation (GEO) when the target is AI answer engines specifically.

The overlap is not a coincidence — both systems solve the same problem: "given this content, which passage answers the question cleanly enough to quote?" A page with vague headings and answers that only make sense with surrounding context loses to both a featured snippet and an AI citation, for the same underlying reason.

Because this is a distinct, growing surface with its own failure modes — llms.txt, crawler access, answerability across a whole site — it is worth auditing separately. Our AI search visibility checker reviews a site against exactly this: whether content is structured for extraction, whether AI crawlers can reach it, and where a page is answerable but not actually being surfaced. Two free tools cover the basics in seconds: check whether AI crawlers like GPTBot are actually allowed with the AI crawler access checker, and create a starter file for them with the llms.txt generator.

Rich results: where structured data comes in

Everything above wins snippets through content structure alone. Rich results are a different mechanism entirely: Google reads Schema.org markup — usually JSON-LD embedded in the page — and uses it to decide whether to render visual enhancements in the listing. Common types worth having, where relevant:

• FAQPage — expandable question/answer dropdowns directly in the search result.
• Product — price, availability, and review-star ratings.
• BreadcrumbList — a breadcrumb trail shown instead of the raw URL.
• Article — publish date and, on some result types, an image.
• HowTo / Recipe — step cards or recipe details, where applicable.

Google's own structured data documentation lists which types are eligible for which treatment, and each type has required and recommended properties. Missing a required property does not just weaken the rich result; it usually disqualifies the page from it entirely, silently, with no visible error on the page itself.

Putting it together: a simple audit for any page

Before publishing a page you want to win a snippet or rich result on:

1. Does at least one section pose the exact question a searcher would type, as a heading?
2. Does the answer land in the first sentence or two after that heading, with no throat-clearing?
3. Does the format — paragraph, list, or table — match what the query actually needs?
4. Is any list or table marked up as real HTML (<ol>/<ul>/<table>), not just visually formatted text?
5. If the page qualifies for a rich result type (FAQ, product, review, breadcrumb), is the matching Schema.org markup present and complete?

The first four are a content and structure job — free, on any platform. The fifth is a markup job, worth doing wherever it genuinely applies, but additive on top of the structure, never a substitute for it.

Wrap-up

Featured snippets are won by structuring content Google already has — a clear question heading, an immediate concise answer, and the right shape for what the query is asking. Rich results are a separate mechanism entirely, driven by Schema.org markup that has to be added, validated, and kept correct as templates change. Do both, but do not confuse them: no amount of structured data will win a snippet for a page that buries its answer, and no amount of answer-first prose will earn a star rating without a Product schema block behind it. Get the structure right first — it is free, it helps every page on the site, and it is exactly what AI answer engines are now looking for too.