Google Discover: How the Feed Works and How to Get Featured in 2026

Google Discover is the personalized content feed that appears inside the Google app and on the Chrome new-tab page on mobile. It surfaces articles, videos, and pages that match a user’s interests without anyone typing a query. For publishers, that makes it one of the largest sources of mobile traffic on the open web — and one of the hardest channels to predict.

This guide explains what Google Discover is, how the feed actually decides what to show, what the February 2026 Discover core update changed, and the practical steps you can take to make your content eligible and competitive.

What Is Google Discover?

Google Discover is an interest-based content feed that recommends web pages, news stories, and videos to logged-in users on mobile. It runs inside the Google app on Android and iOS, on the Google Search homepage on mobile, and on the new-tab page in Chrome on Android. There is no search box involved. Users open the feed and Google fills it with cards selected from its index, ranked against what it knows about that specific user.

Discover is fundamentally different from organic search. In Search, the user types a query and Google returns the most relevant results. In Discover, the user is passive: Google decides when to surface a piece of content based on signals it has built up over time. That makes Discover a push channel, while Search is a pull channel.

Google Discover vs. Google News

Discover and Google News are often confused. Google News is a structured news product organized into topics, categories, and publishers. Discover is broader and more personal. It can include news, but it also surfaces evergreen explainers, product reviews, recipes, hobby content, sports analysis, and almost anything else Google believes a user wants to read. Sites do not need to be in Google News to show up in Discover.

How Google Discover Works

Discover ranks content using interest signals tied to the user, combined with the same quality systems Google uses for Search. Instead of matching a query to a document, the feed matches a document to a person. Google looks at what the user has searched for, the apps they use, the videos they watch on YouTube, their location and Web & App Activity, and the topics they have followed inside the Google app. It then pulls content from its index that matches those interests and that meets Discover’s quality bar.

Eligibility itself is automatic. Google’s documentation states that content is eligible to appear in Discover if it is indexed and complies with Discover’s content policies. No special tag, schema, or feed submission is required. But eligibility is not the same as appearing. Most pages that meet the baseline will never see Discover impressions because they do not match an active user interest, do not pass the quality threshold, or do not have a strong image.

The signals Discover relies on

Public statements from Google and analysis of Discover Performance reports point to a consistent set of inputs the feed uses to rank content:

• User interest signals. Past searches, topics followed, location, and Web & App Activity build the user’s interest graph.
• Content quality signals. The same helpful, reliable, people-first signals Google uses in Search apply to Discover.
• Freshness and timeliness. Discover heavily favors recent content, especially when a topic is trending.
• Page experience. Mobile-first design, fast load times, and Core Web Vitals affect how aggressively Discover distributes a page.
• Image quality. A large, high-resolution image is one of the strongest single signals because Discover is a visual surface.
• Topical authority. Sites that consistently publish strong material on a topic are more likely to be picked for that topic.

Why Google Discover Matters for Publishers

For publishers and content sites, Discover is now too large to ignore. Industry analyses in 2026 have shown that for many major news and content publishers, Discover delivers more referral traffic than traditional Google Search, with reports placing Discover at roughly two-thirds of Google-driven traffic for some publisher categories. Even outside the news vertical, evergreen pages can get sustained Discover impressions long after publication if they continue to match active interests.

The audience Discover sends is different from the search audience. People who arrive from Discover were not actively looking for anything — Google brought your page to them. They tend to scroll, scan, and bounce more quickly than search visitors, but they also tend to be a better top-of-funnel audience because they were curious enough to tap a card. Treating Discover traffic the same as search traffic in your reporting underestimates both channels.

Where Google Discover Appears

Discover is a mobile-first surface and is not available on desktop in any meaningful way. The feed appears in several places:

• Google app. The home screen of the Google app on Android and iOS shows the Discover feed by default.
• Chrome new-tab page on Android. Opening a new tab in Chrome on Android shows Discover cards below the search bar when the feature is enabled.
• Google Search homepage on mobile. Loading google.com on a mobile browser shows Discover beneath the search box for many users.
• Pixel devices. Swiping right from the home screen on a Pixel device opens Discover directly.

Discover is only shown to users who are signed in to a Google Account and who have Web & App Activity enabled. If a user is signed out or has those signals turned off, Google has very little to personalize against, and Discover effectively does not run for them.

Eligibility Requirements for Google Discover

There is no application process and no submission form. Discover pulls from the regular Google index. To be eligible at all, a page needs to be indexable and to meet Discover’s content policies, which prohibit dangerous content, harassment, hate speech, manipulated media, sexually explicit content, terrorist content, and similar categories. Eligibility is binary: either a page can theoretically appear, or it cannot.

The technical baseline Google asks for

• Indexable, mobile-friendly URLs. Discover only ranks pages already in the Google index, and the feed is mobile-only, so the mobile version of the page is what Google evaluates.
• A high-resolution image. Google recommends images at least 1200 pixels wide. The aspect ratio should be close to 16:9, and the file should have at least 300,000 total pixels.
• max-image-preview:large enabled. Without this robots meta directive (or AMP), Google will not show a large image card, which severely limits Discover performance.
• Clear titles and bylines. Discover prefers pages with visible publication dates, author bylines, and authentic, non-clickbait titles.
• Compliance with content policies. Sensational content, misleading headlines, and policy violations remove content from the feed.

None of the technical baseline guarantees a Discover appearance. It only removes the obvious reasons Google would skip a page.

How to Optimize for Google Discover

Optimizing for Discover is closer to optimizing for a recommendation system than for a search engine. You are not trying to match a keyword. You are trying to make Google confident that surfacing your page to a real user interested in this topic will produce a tap, a long read, and no policy issues.

1. Use a Single Strong Image, Not a Generic One

The image is the most visible part of a Discover card. Use a large, original, editorially relevant image at least 1200 pixels wide. Avoid stock photos that look identical to what dozens of other sites would use, avoid logos as the lead image, and avoid composite graphics with embedded text that becomes unreadable on a small card. Set the image as og:image and make sure max-image-preview:large is enabled site-wide.

2. Write Honest, Specific Headlines

Discover rewards headlines that describe the content accurately and gives them a reason to tap. It punishes clickbait, exaggeration, and curiosity-gap titles that do not match the page. The February 2026 Discover core update specifically targeted sensational and clickbait content. A headline should tell the user what the article is about and what they will learn — not tease them. Stay specific, stay accurate, and resist the urge to manufacture urgency that is not in the article itself.

3. Build Real E-E-A-T on the Page

Discover relies on the same helpful, reliable, people-first content signals as Search. Show experience and expertise visibly: a named author with a bio, credentials linked to a real profile, clear publication and update dates, sources cited inline, and an „About“ page that explains who is behind the site. For YMYL topics like health and finance, this is non-negotiable. Even for evergreen lifestyle content, signed bylines outperform anonymous publishing.

4. Optimize for Mobile Page Experience

Discover is mobile-only, so the mobile version of the page is the version that has to work. Pass Core Web Vitals: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift under 0.1. Reduce intrusive interstitials, keep ad density below your editorial content, and avoid layout shift caused by late-loading creatives. Discover users are scrolling — if your page is slow or jumpy, they leave.

5. Match the Content to a Real Interest

Discover only surfaces pages that map to a topic the user already cares about. That means your content has to map cleanly to a topic too. Cover identifiable subjects — a specific product, a named event, a defined how-to — rather than vague, multi-topic posts. Internal evidence from publishers consistently shows that focused, single-topic articles get more Discover impressions than sprawling roundup posts.

6. Add Structured Data Where Relevant

Discover does not require structured data, but the right schema helps Google understand the page. Article or NewsArticle schema is appropriate for editorial content. VideoObject matters when a video is the main asset. Recipe, Product, and How-To schema reinforce the topic for category content. Schema is a clarification, not a ranking factor, but Discover’s quality systems lean on Search systems that do use it.

7. Publish Consistently in a Defined Niche

Sites that publish on a clear set of topics build topical authority that Discover rewards. Consistent cadence helps too: Google’s models look at recent publishing patterns when deciding which sites to test in the feed. Erratic, off-topic posting confuses both readers and the recommendation system.

The February 2026 Google Discover Core Update

On February 5, 2026, Google announced the first core update in Discover’s history that was decoupled from a Search core update. The rollout completed on February 27, 2026, after twenty-one days, initially in English for users in the United States, with international rollout planned for the months after. Google described three explicit goals for the update.

• More locally relevant content. Discover began promoting more content from sites based in the user’s own country and region.
• Less sensationalism and clickbait. Exaggerated headlines, manufactured outrage, and thin content lost Discover visibility.
• More in-depth, original, timely content. Sites with demonstrated, topic-specific expertise gained ground, evaluated on a topic-by-topic basis rather than as one site-wide score.

Independent analyses found measurable state-level personalization after the update — local domains appearing several times more often in their home state’s feed than in another state’s — and a clear shift toward original reporting over aggregator coverage. For most publishers, the practical message of the update is that Discover is now closer to „the best, most original page on this topic from a site that knows this topic“ than it was a year earlier.

How to Track Google Discover Performance

Discover performance is reported in Google Search Console. The Discover report appears as a separate item in the left navigation, but only after a property has accumulated enough Discover impressions to clear Google’s minimum threshold. New sites and small sites often see no Discover report at all.

The report shows clicks, impressions, and click-through rate by URL over the last sixteen months. There is no keyword data because Discover does not run on keywords. Comparing the report against your editorial calendar tells you which formats, topics, and image styles get picked up by the feed and which never break in. Pages that never appear in the report despite strong Search performance are usually missing on image quality, mobile experience, or topical match — not on indexing.

Common Google Discover Mistakes

Most Discover problems trace back to a small number of recurring issues. Treating these as a checklist before publication catches the majority of them:

• Missing or weak hero image. The single most common reason a page never appears. The image must be at least 1200px wide, original, and editorially relevant.
• max-image-preview:large disabled. Without it, Google cannot render the large card layout and Discover effectively skips the page.
• Clickbait or curiosity-gap headlines. The February 2026 update accelerated demotion for these titles. They also rarely return after the update.
• Slow mobile page experience. Heavy ads above the fold, layout shift, and slow LCP all reduce Discover distribution.
• No author or date metadata. Discover prefers pages with visible bylines and publication dates. Anonymous, undated pages get less trust.
• Off-topic publishing. Sites that suddenly cover topics far outside their normal niche rarely break into Discover for those topics.
• Confusing eligibility with appearance. Being eligible is the floor, not the goal. Most eligible pages still never appear.

Frequently Asked Questions

How long does it take to appear in Google Discover?

There is no fixed timeline. New articles can appear within hours of publication if Google’s systems pick up an interest match, but many pages never appear at all. Sites without prior Discover visibility usually need consistent publishing on a defined topic for several weeks before the feed starts testing them.

Can evergreen content appear in Google Discover?

Yes. Discover favors recent content, but evergreen pages can be surfaced when the underlying interest spikes — for example, a how-to guide that resurfaces every time the topic trends. Strong evergreen articles can produce Discover traffic for years.

Does Google Discover work on desktop?

No. Discover is a mobile surface. It runs in the Google app, Chrome on Android, and the mobile google.com homepage. There is no desktop equivalent.

Do I need to be in Google News to appear in Discover?

No. Discover pulls from the regular Google index. Inclusion in Google News is a separate eligibility process and is not required.

Why did my Discover traffic drop suddenly?

Discover traffic is naturally volatile because user interests, trending topics, and Google’s ranking systems all change. Large drops typically come from one of three causes: a Discover or Search core update, a content policy issue (visible in Search Console under Manual Actions), or a technical regression like a broken max-image-preview setting or sitewide image change.

Can I opt out of Google Discover?

Yes. Adding max-image-preview:none in the robots meta tag or using nosnippet removes large-image eligibility. Setting noindex removes the page from Google entirely, which also removes it from Discover.

The Bottom Line

Google Discover is no longer an experimental feed. For many publishers in 2026 it is the largest single source of Google-driven traffic, and the February 2026 core update made clear that Google is willing to tune Discover separately from Search to reward original reporting, local expertise, and honest headlines. Eligibility is automatic, but appearance is not. Pages that earn Discover traffic combine a strong, original image, a precise and accurate headline, real author and topical authority, fast mobile delivery, and tight topical focus — built for a reader who never typed a query and only has to swipe to move on.