TL;DR: We've won 12 featured snippets in the last 6 months. Each one followed the same pattern: exact question as the heading, direct answer in 40-50 words immediately below, proper HTML structure. Featured snippets still drive 35% of clicks when they appear, but they're shrinking on broad queries thanks to AI Overviews. Here's exactly how to optimize for each snippet type — and when to stop chasing them.
Featured snippets are the boxed answers that appear at the top of Google search results — position zero. Google pulls content from a ranking page and displays it directly in the SERP, answering the user's question without requiring a click.
There are four main types: paragraph snippets (the most common at 70%), list snippets (19%), table snippets (6%), and video snippets (5%). Each requires a different optimization approach, and I'll walk through the specific formatting for each one.
The key thing to understand: Google doesn't create snippet content. It extracts it from a page that already ranks on page one. You can't win a snippet if you're not ranking in the top 10 first. Think of snippets as a promotion from existing rankings, not a shortcut to visibility. This distinction matters because I regularly see people trying to "optimize for featured snippets" on pages that rank on page 3. That's not how it works.
I need to be straightforward about where things stand, because the narrative around featured snippets has gotten polarized — either "snippets are dead" or "snippets are essential." Neither is accurate.
Featured snippet visibility dropped 64% between January and June 2025, falling from 15.41% to 5.53% of SERPs, according to multiple rank tracking tools. AI Overviews are eating into snippet territory — Google rarely shows both on the same SERP.
But here's what the panicking crowd misses: snippets are shrinking for broad queries. For specific, fact-based questions — "what is a canonical tag," "how to add schema markup," "average domain authority by industry" — snippets are holding steady. They still capture 35% of clicks when they appear, and they appear on roughly 5-6% of all queries. On a search engine processing 8.5 billion queries per day, that's still hundreds of millions of snippet opportunities.
At SEOJuice, we've won 12 featured snippets in the last 6 months. I track them weekly. Every single one is for a specific, technical query — not a broad informational one. That's not a coincidence; it's where we focus our snippet optimization effort. (I should note that we've also lost 4 snippets during that same period, mostly to pages with fresher data. Snippet retention requires ongoing attention.)
Key Takeaway
Don't optimize for snippets on every page. Target specific, fact-based queries where snippets still dominate. Broad informational queries are increasingly served by AI Overviews instead.
Different query types trigger different snippet formats. Here's the data on what works, and for each format I'll explain the pattern that's worked for our own snippet wins:
| Query Type | Best Format | Example Query | % of Snippets |
|---|---|---|---|
| Definition / "What is" | Paragraph | "what is a 301 redirect" | 70% |
| Process / "How to" | Ordered List | "how to submit a sitemap" | 19% |
| Comparison / Data | Table | "SEO tool pricing comparison" | 6% |
| Tutorial / Visual | Video | "how to use Google Search Console" | 5% |
Paragraph snippets answer "what is" and "does/can/should" questions. Google typically pulls 40-50 words — the sweet spot is 45 words or 293 characters. If your answer is longer, Google truncates it.
The formula we follow:
The key word is immediately. Answer the question in the first sentence after the heading. Don't warm up with "In this section, we'll explore..." or "Many people wonder about..." — just answer. I've watched content that ranks at position 2 lose the snippet to position 5 because the position-5 page answered the question faster. Google rewards directness.
"Answer the question in 40-60 words immediately, not after three paragraphs of throat-clearing. With the flooded market of AI-generated content, Google's 2026 algorithms are hyper-focused on EEAT — to win a snippet, your content must prove it comes from a real expert."
Queries starting with "Does" (99.91%), "Are" (99.86%), "Is" (99.66%), and "Can" (99.74%) are the top generators of paragraph snippets. If your content answers these question types, you're in the best position to win. These percentages come from large-scale SERP analysis data — I include them because they shaped which queries we prioritized.
"How to" queries trigger list snippets 46.91% of the time. Google pulls content from either HTML lists (<ol>, <ul>) or heading-based structure (a series of H2s or H3s).
The HTML structure matters more than most people realize. We tested this directly: we had a page ranking at position 3 for a "how to" query with the steps written as paragraphs. We reformatted them into an ordered list. The snippet appeared within 10 days. No other changes.
<!-- Ordered list for step-by-step (Google extracts this directly) -->
<h2>How to Submit a Sitemap to Google</h2>
<ol>
<li>Log in to Google Search Console</li>
<li>Select your property from the dashboard</li>
<li>Click "Sitemaps" in the left sidebar</li>
<li>Enter your sitemap URL (e.g., /sitemap.xml)</li>
<li>Click "Submit" and verify the status</li>
</ol>
<!-- Heading-based list (Google assembles this from H3s) -->
<h2>Top SEO Mistakes to Avoid</h2>
<h3>1. Missing title tags</h3>
<p>Details about the mistake...</p>
<h3>2. Broken internal links</h3>
<p>Details about the mistake...</p>
<h3>3. No schema markup</h3>
<p>Details about the mistake...</p>Something we've observed: Google displays up to 8 items in a list snippet, then adds "More items..." with a link to your page. If your list has exactly 8 or more steps, this creates a curiosity gap that drives clicks. Three of our snippet wins use this pattern deliberately.
Table snippets appear for comparison and data queries. Google extracts them directly from <table> HTML elements — and only from proper table elements. CSS grids, flexbox layouts, and div-based "tables" don't work. I've tested this specifically because we build a lot of comparison pages.
The optimal table size for snippet extraction: 3 columns, 5-6 rows. Keep it tight. Google won't display a 15-column monster.
Key Takeaway
Use actual <table> HTML, not CSS grids or divs styled to look like tables. Google cannot extract structured data from non-table elements for snippet display.
Video snippets pull from YouTube almost exclusively. Google shows a video thumbnail with a specific timestamp that answers the query. We don't have direct experience winning video snippets since we don't produce video content yet, but the documented best practices are consistent across sources:
0:00 - Introduction, 1:23 - Step 1)This is the tactical part, and it's where most of our 12 snippet wins came from. We weren't creating content for new queries — we were taking snippets away from competitors who owned them. Here's the process:
Search the query. Look at who owns position zero. Open their page and find the exact content Google is pulling. Note the format (paragraph, list, table).
Look at their heading, the answer paragraph length, and the surrounding content. Note what's missing — is the answer outdated? Is it too vague? Does it lack specific data?
Match their format exactly but make your answer:
Does their page only answer the primary question? Add a table, a list, and related FAQs. Google's snippet selection considers the overall page quality, not just the snippet-eligible paragraph. This is where we've seen the most consistent wins — adding a comparison table to a page that previously only had paragraphs.
Internal links from high-authority pages on your site. Fresh backlinks if possible. Schema markup (FAQ schema on the same page can trigger dual snippet + PAA appearances).
"Data-backed snippets show 40% higher citation rates than purely conceptual answers. AI models prioritize verifiable information — statistics, dates, specific numbers, percentages, and concrete examples."
Winning a snippet means nothing if you don't know you won it — or if you lose it three days later. We lost 4 snippets in the last 6 months, and in two cases I didn't notice for over a week because we weren't monitoring closely enough. Here's the system we use now:
| Metric | What It Tells You | Check Frequency |
|---|---|---|
| Snippet ownership | Whether you hold the snippet for target queries | Weekly |
| CTR at position 1 | Whether the snippet is driving clicks (not just impressions) | Monthly |
| AI Overview presence | Whether an AI Overview has replaced the snippet SERP | Monthly |
| Competitor snippet changes | Who's winning/losing snippets in your niche | Bi-weekly |
| Snippet format shifts | Whether Google changed the format (paragraph to list, etc.) | Monthly |
In Google Search Console, you can spot snippet holders by filtering by query and looking for pages with unusually high CTR at positions 1-3. A page with 40%+ CTR at position 1 is likely a snippet holder. Our SERP features tool flags snippet ownership directly, which saves the detective work.
I mentioned we lost 4 snippets in 6 months. Here is what happened with each one, because losses teach you more than wins:
Loss 1: "What is crawl budget" — lost to fresher data. We held this snippet for about four months. A competitor published a new article that referenced Google's 2026 crawl documentation updates and included specific numbers from the March 2026 core update. Our article still referenced 2024 data. Google swapped us out within two weeks of their publication. Lesson: snippet retention requires content freshness. We now review our snippet-holding pages monthly for stale data points.
Loss 2: "How to fix broken links" — lost to a better format. This one stung. We had a paragraph snippet. A competitor restructured their answer as a numbered list with the exact same information. Google switched the snippet format from paragraph to list and gave it to them. Same content, different HTML structure. We reformatted our answer into an ordered list and won it back three weeks later. Lesson: when the query implies steps, use <ol>, not paragraphs.
Loss 3: "SEO audit checklist" — replaced by AI Overview. Google started showing an AI Overview for this query, and the featured snippet disappeared entirely. No one holds it now. There is nothing to win back. We redirected our optimization effort to more specific sub-queries ("technical SEO audit checklist", "on-page SEO checklist") where snippets still appear. Lesson: before investing in snippet optimization, check if the query already triggers an AI Overview. If it does, move on.
Loss 4: "What is domain authority" — lost to Moz. Moz literally invented the term. They updated their definition page, and Google gave the snippet to the canonical source. We held it briefly because our answer was more concise, but once Moz tightened their definition to snippet-length, there was no competing. Lesson: for branded or creator-owned terms, the originator has an inherent advantage. Focus your snippet efforts on neutral queries where no single source owns the concept.
Across our 12 wins and 4 losses, the clearest pattern was about format matching. We tracked which HTML format Google selected for each snippet and compared it to the query type:
<ol>. When we initially had these as paragraph descriptions of steps, we did not win the snippet. Reformatting to lists triggered the win.<table> with 3 columns and 4-5 rows. We tested the same content as bullet lists first — no snippet. Changed to a table — snippet appeared within two weeks.This is not revolutionary insight, but seeing it play out in our own data made it feel less like "SEO advice" and more like a mechanical rule. Match the format to the query intent. Paragraph for definitions. List for processes. Table for comparisons. The content quality matters, but if the format is wrong, you will not win regardless of how good your answer is.
This is the elephant in the room, and I'm not going to pretend it doesn't matter.
AI Overviews now appear in an estimated 47% of searches. They reduce clicks to top-ranking content by 34-58% depending on the study and the query type. Featured snippets and AI Overviews rarely appear together — Google treats them as mutually exclusive.
| SERP Feature | CTR Impact | SERP Presence | Trend |
|---|---|---|---|
| Featured Snippet (no AI Overview) | +35% CTR | ~5.5% | Declining |
| AI Overview (no snippet) | -58% CTR | ~47% | Growing |
| Neither (traditional SERP) | Baseline | ~47.5% | Stable |
So what do you actually do? Here's how I think about it:
The good news: the same content principles work for both. Clear structure, direct answers, data-backed claims, and proper HTML formatting. If you optimize for snippets, you're also building content that AI Overviews are likely to cite. We've seen this firsthand — two of our snippet-optimized pages are also cited in AI Overviews for related queries.
Yes, but the opportunity is more targeted than before. When a snippet appears without an AI Overview, it captures roughly 35% of clicks. Focus on specific, fact-based queries where snippets still dominate. We continue to invest effort in snippet optimization, but only for the right query types.
Extremely unlikely. Google almost always pulls snippets from pages ranking in the top 10. Get to page 1 first, then optimize for the snippet. In our experience, trying to win a snippet from beyond position 10 has never worked.
FAQ schema doesn't directly win featured snippets, but it can get your content into People Also Ask boxes, which function similarly. More importantly, FAQ schema signals to Google that your content is structured for question-answer extraction — which correlates with snippet eligibility.
We've seen snippet wins appear within 10 days of a content restructure, but typically it takes 2-4 weeks. If you're already on page 1, reformatting your content for snippet eligibility can produce results within a single crawl cycle.
This is a real concern. Some snippets answer the question so completely that users don't click through. If you're seeing high impressions but low CTR on a snippet query, consider making your snippet answer slightly less complete — include enough to earn the snippet, but add a hook that requires clicking for the full answer. We've done this on two pages with measurable CTR improvement.
Featured snippets are smaller than they were in 2023, but they're not dead. They're more competitive, more selective, and more dependent on content quality than ever. The sites winning snippets in 2026 — including ours — aren't using tricks. They're answering questions better than anyone else on page 1, with tighter formatting and more specific data.
Format your content for machine extraction. Answer questions in 45 words. Use real HTML tables and lists. Back your claims with data. And check whether your target query even shows a snippet before spending time optimizing for one. That last step saves us hours of wasted effort every month.
no credit card required