Rich Features / Discover / Images / Video / Page Experience / Web Stories / SXG / Paywall Complete Guide

One sentence: This wiki covers Google's "non-SD-dominated" rich presentation features — Discover / Images / Video / Page Experience / Web Stories / SXG / Paywall / Site Names / Reviews system / Package Tracking / etc.

TL;DR — Which feature do you want to implement?

Want to implement	See
Appear in Google Discover feed	Part 1 — No special tag, based on interest + large image + page experience
Images indexed by Google Images + images shown in text results	Part 2 — `<img src>`, image sitemap, og:image, SafeSearch
Videos appear in video results / Key Moments / Live	Part 3 — watch page + video SD + allow fetch
Page Experience optimization / Core Web Vitals	Part 4 — LCP/INP/CLS
Avoid intrusive interstitials harming SEO	Part 5 — Use banner instead of modal
SXG lets Search prefetch speed up performance	Part 6 — Web Packager / AMP
Show your site name in Search results	Part 7 — WebSite SD + alternateName
Create Web Stories	Part 8 — valid AMP + self-referential canonical + 5 content policies
News paywall but Google can index	Part 9 — metering + lead-in + paywall SD
Logistics companies show package tracking in Search	Part 10 — Early participation, India/Japan/Brazil
Display article published / updated date	Part 11 — visible date + Article SD
Restaurants appear in "Top X" lists	Part 12 — Citing site compliance
Make review sites rank well	Part 13 — reviews system

Part 1 — Google Discover

1.1 What it is + triggers

Discover = Google Search's interest feed, which automatically surfaces relevant content based on user Web/App Activity. Primarily mobile, iOS / Android.

1.2 How to get into Discover

Fully automatic,No special tag / no SD required. Discover uses the same ranking signals as Search and systems.

Eligibility:

Indexed by Google + meets Search technical requirements
Meets Discover content policies

⚠️ Eligible Does not guarantee appearance.

1.3 Improve appearance probability (7 best practices)

Avoid clickbait — Titles / previews / images must not be misleading / exaggerated
Page title captures the core — Headline clearly states the essence of the content
Avoid sensationalism(morbid / titillation / outrage types)
Timely content / good stories / unique insights
Large image is mandatory:
- ≥ 1200px wide
- Total pixels > 300,000(16:9 1280x720 = 921,600 ✅)
- 16:9 aspect ratio recommended
- Must be <meta robots content="max-image-preview:large"> or AMP
Use primaryImageOfPage SD or og:image to specify preferred image (do not use logos / text-heavy images)
good page experience(see Part 4)

1.4 Discover Content that may be filtered content

Job applications / petitions / forms
Code repositories
Satirical content without context
Content filtered by SafeSearch (explicit)
"shocking / unexpected" content

1.5 Why Discover traffic changes

Discover is supplemental traffic,Do not count it as primary traffic.

3 reasons for fluctuation:

User interest shifts(users no longer search for the topic → Discover stops recommending it)
Discover adjusts content types(platform-level content reselection)
Search updates affect Discover(because it is a Search extension)

1.6 Monitoring

Search Console Discover Performance report(past 16 months, meets impression threshold) — includes Discover traffic within Chrome.

Part 2 — Google Image SEO

2.1 Let Google discover your images

A. Use HTML <img> element(mandatory):

✅ <img src="puppy.jpg" alt="A golden retriever puppy" />
❌ <div style="background-image:url(puppy.jpg)">...</div>  <!-- CSS 图 Google 不索引 -->

Supported src in <img> or <picture> (including nested). Data URI is also OK but increases page size so use with caution.

B. Image sitemap(optional) — image sitemap allows <image:loc> cross-domain (used for CDNs).

C. Responsive images(<picture> / srcset) — always include <img src> fallback.

2.2 Supported formats

BMP / GIF / JPEG / PNG / WebP / SVG / AVIF(new in 2026)

2.3 Specify preferred image

3 methods (choose one):

A. schema.org primaryImageOfPage:

{
  "@context": "https://schema.org",
  "@type": "WebPage",
  "url": "https://example.com/url",
  "primaryImageOfPage": "https://example.com/images/cat.png"
}

B. image within mainEntity / mainEntityOfPage:

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "mainEntityOfPage": "https://example.com/url",
  "image": "https://example.com/images/cat.png"
}

C. og:image meta tag:

<meta property="og:image" content="https://example.com/images/cat.png">

2.4 4 image selection guidelines

Relevant to the page + representative
Avoid generic images(logos) / text-heavy images
Avoid extreme aspect ratios (too narrow / too wide)
High resolution(especially Discover 1200px+)

2.5 Alt text + filename (5 guidelines)

❌ 缺 alt:               <img src="puppy.jpg"/>
❌ keyword stuffing:    alt="puppy dog baby dog pup pups doggies..."
⭕ Better:               alt="puppy"
✅ Best:                 alt="Dalmatian puppy playing fetch"

filename descriptive:my-new-black-kitten.jpg is not IMG00023.JPG
Large sites can use programmatic naming
When localizing, translate filenames as well (URL-encoded)
SVG uses <title> + aria-labelledby

2.6 Opt out of inline linking (do not show original image in Google Images, only show thumbnail)

HTTP referrer = google domain → 返 200 / 204 空响应

→ Google still crawls + indexes + shows thumbnails,this is not cloaking and will not be penalized.

Or completely: go through prevent-images-on-your-page.

Part 3 — Video SEO

3.1 6-step process

1. 帮 Google 找视频(标准 HTML 元素 + 不要 fragment / 不要 user 触发)
2. 确保可索引(watch page 索引 + 单视频 dedicated 页 + 合法 thumbnail + stable URL)
3. 开启 specific feature(video previews / Key Moments / Live Badge)
4. 必要时移除 / 限制
5. Search Console 监控
6. Troubleshoot

3.2 Indexing requirements


Watch page must be indexable	Dedicated page for a single video
Watch page must perform well in Search	Being indexed alone is not enough; performance is required for the video to be indexed
Video is not blocked by other elements	Paywalls use paywalled-content SD
valid thumbnail + stable URL	At least 60x30px, 80%+ alpha > 250 (main subject opaque)

3.3 watch page vs non-watch page

Is a watch page:

Video landing page
TV single episode page
News video watch page
Sports highlight page
Event clip page

Is not a watch page(video is supplementary):

Blog post (review embed)
Product page (360 video)
Video category page (lists multiple videos)
Movie review page (embedded trailer)

→ Non-watch pages can still appear as text results + Google Image with video badge but do not get video-specific features.

3.4 3 key URLs

URL	Tag location
Watch page	sitemap `<loc>`
Video player	`<iframe src=>`;SD `embedUrl` or sitemap `<video:player_loc>`
Video file(media bytes)	`<source src>` / `<embed src>` / `<object data>`;SD `contentUrl` or sitemap `<video:content_loc>`

3.5 3 major video features

A. Video Previews

Google extracts a few seconds of moving preview
Must allow Google to fetch video files
max-video-preview:[秒] robots meta controls duration

B. Key Moments (chapters)

2 paths:

Clip SD — You specify start/end + label (all languages)
SeekToAction SD — You tell Google how the URL contains timestamps, Google finds them automatically (English/Spanish/Portuguese/Italian/Chinese/French/Japanese/German/Turkish/Korean/Dutch/Russian)
YouTube — Add timestamp + label in description (all languages)

→ nosnippet opt out.

C. Live Badge

Livestreams use BroadcastEvent SD.

3.6 Must allow Google to fetch video files (M3U8, etc.)

Do not noindex / robots.txt block video byte URLs
Stable URL
contentURL Use supported formats
Both watch page host + streaming server meet technical requirements + capacity

Worried about exposure?→ verify Googlebot Only expose to real Googlebot contentUrl.

3.7 Remove videos

Method	Applicable scenario
Removals request	Urgent — watch page must be truly inaccessible (404 / noindex / auth) otherwise it will be reindexed
Return 404	Permanent deletion
noindex robots meta	Do not index watch page
`<video:expiration_date>` / SD `expires`	Do not show video results after expiration (but watch page can still appear as a text result)

⚠️ To remove third-party embedded videos you need to contact the other party |

3.8 Country restrictions

Method	Configure
SD `regionsAllowed` / `ineligibleRegion`	allow list / deny list
sitemap `<video:restriction relationship="allow">CA MX</video:restriction>`	ISO 3166-1 two- or three-letter codes

Part 4 — Core Web Vitals + Page Experience

4.1 Core Web Vitals 3 metrics (all must be good)

Metric	Meaning	Target
LCP(Largest Contentful Paint)	Loading performance	< 2.5s
INP(Interaction to Next Paint)	Interaction responsiveness	< 200ms
CLS(Cumulative Layout Shift)	Visual stability	< 0.1

Measurement tools:

Search Console Core Web Vitals report
PageSpeed Insights
web.dev vitals tools

4.2 Page Experience is more than CWV (6-dimension self-assessment)

Good Core Web Vitals
HTTPS security
mobile-friendly
No excessive intrusive ads
No intrusive interstitials(see Part 5)
Main content Easily distinguishable from other content

4.3 Page Experience FAQ

Q: A single ranking signal? A: No. Multiple signals are weighted.

Q: Which page experience factors actually affect rankings? A: Core Web Vitals are actually used. Others do not directly affect rankings, but they make the site easier to use → indirect help.

Q: Evaluated at site level or page level? A: Mostly page-level, with a small number of site-wide evaluations.

Q: How important is it? A: When multiple pages have similarly helpful content,page experience acts as a tiebreaker. But Google still prioritizes relevance.

Part 5 — Avoid Intrusive Interstitials

❌ Full-screen interstitials that block content ✅ Top / bottom banners that occupy a small portion of the screen

For App install prompts use:

iOS: Smart App Banners
Chrome: in-app install experience

For Newsletter signups use a CMS plugin.

5.2 Two major mistakes you must avoid

❌ Covering the full page interstitial
❌ Redirecting users to a separate consent page

5.3 Mandatory Interstitials (legally required)

For legally required cases such as casino / adult / GDPR consent:

Use an interstitial overlay for content(so that Googlebot can index at least some content)
❌ Do not redirect all URLs to a single consent page(otherwise all pages will be removed from the index)

Adult content: verify Googlebot → For the real Googlebot skip the age gate directly.

Part 6 — Signed Exchange (SXG)

6.1 What it is

Google Search Prefetches your site resources (HTML/JS/CSS/img/font), so after the user clicks it opens almost instantly. LCP decreases → page experience score improves.

Supports both AMP and non-AMP.

6.2 Implementation

web.dev SXG in-depth + Chrome LCP optimization

AMP: amp.dev SXG guide

6.3 expiration values rules

Cannot exceed the HTTP header cache expiration
JS or inlined JS content → < 1 day
Others → < 7 days

6.4 Multi-device adaptation

Place personalized content (shopping cart, etc.) outside lazy-loaded elements outside of SXG
or Vary: Cookie signed header (only for cookieless users)
Responsive web design or separate URL + supported-media meta:

<meta name=supported-media content="only screen and (max-width: 640px)">

6.5 Debugging

SXG Validator Chrome extension
Check the Google SXG cache directly:
- SXG URL https://signed-exchange-testing.dev/sxgs/valid.html
- cache URL https://signed--exchange--testing-dev.webpkgcache.com/doc/-/s/signed-exchange-testing.dev/sxgs/valid.html
- The algorithm is the same as the AMP cache,infix /doc/-/ different
Warning header = SXG does not meet cache requirements
Location header = cache has not fetched yet

Part 7 — Site Names

7.1 What it is

In Search results the site name(independent of the title link). Example: "Burnt Toast" → title "How to make toast in a pan".

7.2 Implementation (WebSite SD)

<script type="application/ld+json">
{
  "@context" : "https://schema.org",
  "@type" : "WebSite",
  "name" : "Example",
  "alternateName" : "EC",  // 备选(数组也可:["BT", "B-T", "Burnt Toast Shop"])
  "url" : "https://example.com/"
}
</script>

Place it on the homepage's <head>.Only the home page needs it, inner pages do not.

7.3 Technical rules

1 site 1 name(domain or subdomain level)
✅ https://example.com, https://www.example.com, https://m.example.com(www/m are equivalent)
✅ https://news.example.com(subdomain home)
❌ https://example.com/news(subdirectories are not supported)
Must be on the home page crawlable
For multiple versions of the home page (http vs https) use the same SD

7.4 Five naming guidelines

unique Accurately reflect the identity
Commonly-recognized and concise("Google" not "Google, Inc")
Avoid generic names("Best Dentists In Iowa")
Consistent with other text on the homepage(<title>/og:site_name/heading)
alternateName as a backup(strongly considered when Google does not select the preferred name)

7.5 What if it is not selected?

1. 核 WebSite SD 准确 / 无 syntax 错
2. 其他 source 一致
3. 不是 subdirectory
4. redirect 工作正常
5. 改完等 recrawl

仍不行 → 4 级 fallback:
  L1: name 写首选
  L2: 加 alternateName(强烈考虑)
  L3: alternateName 含**全小写域名**作 backup("example.com")
  L4: name 直接写全小写域名(最后手段)

Part 8 — Web Stories

8.1 What it is + where it appears

Short video / image / text Story format (similar to IG / Snapchat Stories).The underlying web page,must be valid AMP.

Can appear in:

Individual Google Search results (globally)
Individual Discover cards (globally,most common in US/India/Brazil)

8.2 Five steps to launch

1. 创建(无代码用 Story editor / 有工程用 amp.dev)
2. 保证 valid AMP(用 AMP Test Tool / AMP Linter)
3. 验证 metadata(4 必填:publisher-logo-src / poster-portrait-src / title / publisher)
4. 检查是否被索引(URL Inspection / Page Indexing / Sitemaps report)
5. 遵 5 内容政策

8.3 Five content policies (violation → loss of richer experience)

No copyright infringement
Not text-heavy(≤ 180 characters per page; videos <60s encouraged)
No low-quality assets(blurry / pixelated)
Must have a narrative arc(coherent theme)
Not incomplete(do not require users to click external links for key information)
Not overly commercial(a small amount of affiliate content is allowed, cannot be pure advertising)

8.4 Critical Best Practice (must do)

| Storytelling | Video first | | Design | ≤ 280 characters per page + Text is not covered + Text does not overflow + Use animations sparingly | | SEO | High-quality content + title < 90 characters + Do not use noindex + Self-referential canonical + AMP metadata | | Technical | valid AMP + Poster contains no text + poster ≥ 640x853 + 3:4 + logo ≥ 96x96 + 1:1 |

8.5 Recommended Best Practice (as much as possible)

Video < 15s / page (60s maximum)
High-quality audio ≥ 5s
For video-only Stories consider auto-advance
Add desktop landscape support (to appear on desktop)
Include structured data + alt text + captions
Integrate Web Stories into the main site URL structure (no need to mark "amp" / "stories" in the URL)

Part 9 — Flexible Sampling(Paywall)

9.1 Two types of sampling

Method	Meaning
Metering	Give users a monthly quota(monthly recommended, not daily). Start at 10 articles/month,6–10 articles is a reasonable range for news sites
Lead-in	Above the paywall show the first few sentences of the article(increases curiosity + lets Google index some content)

9.2 Key threshold

The threshold at which Google user experience drops significantly: paywall trigger rate > 10%(about 3% of audiences see the paywall).Avoid getting close to this line.

Metering should apply monthly rather than daily, because "10 → 8" has less impact than "3 → 2", and it focuses the paywall on the most active users.

9.3 Mark paywall content

Must use paywalled-content structured data, otherwise it is treated as cloaking — Googlebot sees the content but users are blocked.

If you do not want to give content to the browser → use an implementation that does not serve paywalled content to the browser implementation.

Part 10 — Package Tracking (early participation)

10.1 Eligibility

Logistics companies based in India / Japan / Brazil(or other companies that are exclusive package tracking providers in those regions)
Must have a REST JSON API(POST requests)
Response time 700ms average, 95th percentile < 1000ms

10.2 API requirements

Required fields: CurrentStatus(including timestamps and error statuses)

Recommended fields: DeliveredDate / PromisedDate / TrackingNumber / TrackingURL / SupportPhoneNumbers / TransitEvents / CreateDate / PickupDate / TimestampEvent / LocationEvent / CanReschedule

Not allowed: recipient / sender personal information / geographic information.

10.3 Sign up

Early Adopters Program form

Part 11 — Byline Date (publication date)

11.1 Implementation (3 steps)

Add user-visible date in a prominent position ("Posted Feb 4, 2019" / "Last updated: Feb 14, 2019 8pm ET")
Add to SD datePublished and/or dateModified(use Article / BlogPosting / NewsArticle / VideoObject and other CreativeWork subclasses)
Visible content + SD must be consistent

{
  "@context": "https://schema.org",
  "@type": "NewsArticle",
  "headline": "Analyzing Google Search traffic drops",
  "datePublished": "2021-07-20T08:00:00+08:00",
  "dateModified": "2021-07-20T09:20:00+08:00"
}

11.2 6 best practice

Date is required, time is optional(but adding time + timezone improves precision)
Correct timezone (consider daylight saving)
Visible content + SD consistent
Do not write future dates / event dates(it should be the publication/update date)
Reduce other dates on the page interference
News pages → News additional guidelines

Part 12 — Top Places List

12.1 What it is

If your store is on some "Top X restaurants" list, Search results may surface these lists.Brick-and-mortar businesses only.

12.2 Three compliance conditions for list sites

Curated by the content provider, truly independent,not sponsored
Not automatically generated from templates(data / automatic metrics are not acceptable)
No vulgar / offensive language

12.3 I don't want to appear?

→ Opt out of specific Google properties in local search results.

Part 13 — Reviews System

13.1 What it is

Google's dedicated evaluation system for review content, which rewards reviews with deep research / written by experts review,and demotes thin summaries.

13.2 Evaluation scope

First-party articles / blogs / pages that provide recommendation / opinion / analysis of
Does not evaluate third-party reviews (user comment sections)
Single-product / comparison / ranked lists all count

13.3 Impact scope

Primarily page-level evaluation
If a site has a large amount of review content → site-level evaluation is possible(including non-review content)

13.4 Supported languages (11)

English / Spanish / German / French / Italian / Vietnamese / Indonesian / Russian / Dutch / Portuguese / Polish

13.5 Improve content quality

→ For the full 13 guidelines see Complete Guide to Ecommerce SEO (6 Surfaces / 4 Launch Strategies / Merchant Center / 6 SD Types / Pagination 3 Modes / Review 13 Guidelines) Part 8.

In short:

Expert perspective
Original research + quantitative metrics
Compare with competitors + evolution path
First-hand evidence (screenshots / videos)
Avoid thin affiliate content

Product reviews can use Product SD to help Google identify them.

13.6 How long to recover after changes?

Reviews evaluation is automated,and works together with other ranking factors. Recovery is possible after changes,but timing is not guaranteed.

Part 14 — Related wiki

Search result UI elements (title/snippet/favicon/sitelinks/featured): Complete Visual Guide to Search Result Elements (Visual Gallery / Title / Snippet / Favicon / Sitelinks / Featured / AI / Translated / Preferred)
Complete Structured Data type list: Complete Overview of Structured Data (Framework + 34 Rich Result Types for Quick Lookup + JSON-LD/Microdata/RDFa Selection)(upcoming wiki)
robots meta(nosnippet/max-snippet/notranslate): Complete Guide to Crawling and Indexing Control (robots.txt / noindex / X-Robots-Tag / data-nosnippet / rel)
canonical + AMP relationship: Complete Guide to Canonical / Redirects / Site Migration (including full HTTP status code table + 503 temporary closure playbook)
Relationship between Page experience and mobile-first: Complete Guide to Googlebot Crawling Mechanisms and Indexing Control (URL/Links/Mobile-first/Meta/Testing/AMP) Part 5
Video sitemap fields: Complete Guide to Sitemaps (XML / RSS / Text + image / video / news / hreflang extensions + sitemap index) Part 7
Web Stories are a subset of AMP / AMP is in overall decline but Web Stories remain an entry point: Complete Guide to Googlebot Crawling Mechanisms and Indexing Control (URL/Links/Mobile-first/Meta/Testing/AMP) Part 10
13 Guidelines for Writing Reviews: Complete Guide to Ecommerce SEO (6 Surfaces / 4 Launch Strategies / Merchant Center / 6 SD Types / 3 Pagination Modes / 13 Review Guidelines) Part 8
Helpful content + spam policies (relied on by both Discover and Reviews): Google SEO Getting Started + How Search Works + Search Essentials

Last updated

2026-05-17 — Compiled based on Google Search Central's 16-part appearance feature series (latest from 2025-12 to 2026-03). Includes Discover 2026-03 / Image 2026-03 / Web Stories best practices 2026-03.

2026-09 Supplement: Regional search displays and cross-platform identity entry points

This section is based on official documentation snapshots from 2026-09-08 and 09-16. Eligibility is determined separately by region, business identity, and query type. Do not apply the same set of structured data to all pages just because you see one display format.

Regional Display Selection

Feature	Regions and Scenarios	Identity, Data, and Participation Requirements
Aggregator unit	EEA (European Economic Area); hotels, flights, long-distance rail/bus, products	VSS (Vertical Search Services), including OTAs (Online Travel Agencies), price comparison sites, metasearch engines, and directories; requires approval, relevant content, quality compliance, and data access; submit the applicable interest form separately
Supplier unit	EEA; appears alongside the corresponding aggregator unit	Direct suppliers such as hotels and airlines; appears only when an aggregator unit is present; crawlable information on web pages can serve as a source, and additional data sources can enhance results
Ecosystem carousel	EEA desktop and mobile; weather, sports, finance, translation	Authoritative or professional service providers; no new dedicated feed or structured markup required; eligible parties can express interest in participating
Job sites (job aggregation carousel and filters)	EEA; job search queries	Aggregation site entry point; no additional markup required; "More sites" provides access to more aggregated results
Places sites (places aggregation carousel and filters)	Türkiye; hotels and local businesses	Aggregation site entry point; no additional markup required; interest can be expressed via the applicable form
South African (South Africa badge and filters)	South Africa; travel, products, car rentals, buses, food delivery, etc.	Only eligible South African platforms; the badge itself does not change rankings, and filters do not require additional markup; refer to documentation separately for the applicable scope of ads and organic results

Do not confuse the above aggregation entry points with Structured data carousels: the latter form horizontally browsable results from single-site entity lists and require implementation of the corresponding markup. The regional overview lists different query scopes for the EEA, South Africa, and Türkiye; support in one region cannot be inferred to apply identically to other regions.

Actual work required for aggregator onboarding

The unit expands results from the highest-ranked provider by default, and users can select another provider; only one aggregator unit is displayed at a time, and clicks within the unit go directly to the corresponding aggregator website. Obtaining eligibility does not guarantee being expanded by default at all times.

Hotels are onboarded via the applicable accommodation point-of-interest data source, transport and flights via the corresponding real-time APIs, and products via the CSS (Comparison Shopping Service) process. Operational quality focuses on:

Entity descriptions, categories, facilities, opening hours, and trusted ratings should be as complete as possible to help users compare; generic terms must not be used in place of specific entities.
Names should be truthful and descriptive; avoid all caps, promotional language, excessive punctuation, or emojis.
Images should be clear, original, and free of promotional overlays or watermarks; provide high-resolution versions when available.
Prices and availability should match landing pages as closely as possible, and expired or out-of-stock results should be removed promptly. Complete data fields with outdated prices still harm user experience.
Monitor overall performance in Search Console; do not treat "having submitted a feed" alone as evidence of successful display.

Search profile badge: Help users find cross-platform content

A Search profile aggregates a website's associated content from Instagram, TikTok, YouTube, X, Facebook, and other platforms into a single Google entry point. Official guidance notes that following a profile may make that content more likely to appear in that user's Discover, but no fixed recommendation volume is promised.

After claiming the profile, use https://profile.google.com/@handle, and replace handle with the actual account. You may use officially downloaded badge assets or place a plain text link directly:

<a href="https://profile.google.com/@example">Find us on Google Search</a>

The displayed text means "Find us on Google Search"; the example in the example needs to be replaced. If using an icon, provide an accessible name and alternative text. The clickable area must be at least 48×48 dp on Android and 44×44 px on iOS and web per documentation; do not stretch, rotate, or recolor official logos, and keep monochrome and color schemes consistent. When displayed alongside the Preferred Sources badge, the Search profile should use a labeled, higher-emphasis button to avoid confusing users with two similar G icons serving different purposes.