Google Discover: How to Get Your Content Into the Feed

Everything Google documents for Discover, on the card itself: a large image, the large-preview rule, an honest headline, an indexed page and good page experience.

If a "Discover" tab has appeared in your Search Console, your site has crossed a threshold most sites never reach. The question is what to do with it.

This guide is part of our SEO guides hub, and it covers Discover the way Google documents it — not the way it gets mythologised on LinkedIn.

What is Google Discover?

Google's definition is short: Discover is "a part of Google Search that shows people content related to their interests, based on their Web and App Activity" (Google Search Central).

The key word is interests, not queries. In Search, a person types something and Google matches pages to it. In Discover, nobody types anything. Google looks at what a person has been reading and following, and pushes cards it thinks they will want next.

Google's user help page calls it a way to "get updates for your interests, like your favorite sports team or news site, without searching for them." It appears in the Google app, on google.com in a mobile browser on Android, and on some phones by swiping right from the home screen. It is a mobile surface first.

How is Discover different from Search, AI Overviews and featured snippets?

Discover is the only one of the three with no query behind it. That single difference changes how you optimise for it, measure it, and how much you can rely on it.

A featured snippet rewards a clean answer to a known question. An AI Overview rewards being a citable source for a known question. Discover rewards being interesting enough that someone who never asked a question stops scrolling. Top Stories, video carousels and People Also Ask are covered in our SERP features guide — none of them is Discover either.

Who is eligible for Discover?

Everyone whose content is indexed. Google states that content is eligible if it is "indexed by Google" and meets "Discover's content policies," and that "no special tags or structured data are required" (Search Central). No form, no feed submission, no Google News approval.

But Google adds one sentence to read twice: "being eligible to appear in Discover is not a guarantee of appearing." Eligibility is the floor.

What kind of content gets into Discover?

Google's guidance names three things Discover favours: content "that's timely for current interests, tells a story well, or provides unique insights." It also says Discover "makes use of many of the same signals and systems used by Search to determine what is helpful, people-first content."

In practice, that maps to a few content shapes (shapes, not named publishers — we could not verify specific Indian examples this run):

• News and explainers tied to something happening now. A budget announcement, an exam result date, a cricket series. Interests spike and fade, so Discover leans on freshness.
• Evergreen pieces with a strong visual. A well-photographed recipe, a travel itinerary, a before-and-after home story. The image is the card.
• Deep pieces on a topic a reader has already shown interest in. If someone reads three articles about mutual funds, a genuinely useful fourth one from you can land in their feed.

What rarely works: thin roundups, pages whose only image is your logo, and anything built to be clicked rather than read.

The technical checklist Google actually documents

Most "Discover optimisation" advice is guesswork. This list is not — every item is in Google's Discover or page-experience documentation, and each is a fix you can ship this week.

1. Use a large, high-quality image on every article. Google says to "include compelling, high-quality images," that large images "need to be at least 1200 px wide," and that they are "more likely to generate visits from Discover" (Search Central). The same page recommends a 16:9 ratio and says an og:image tag or schema.org markup can point Google at your preferred image.
2. Add the max-image-preview:large robots rule. A big image is useless if Google may only show a thumbnail. Add <meta name="robots" content="max-image-preview:large"> to the page head; Google's robots meta tag documentation defines large as "a larger image preview, up to the width of the viewport." Most WordPress SEO plugins emit this by default — check your page source rather than assuming.
3. Make the page fast on a phone. Google's page-experience guidance asks whether your pages "have good Core Web Vitals" and states they "are used by our ranking systems" (Search Central). Our Core Web Vitals guide covers the three metrics and how to fix each.
4. Remove intrusive interstitials. The same guidance asks whether your pages "avoid using intrusive interstitials." A full-screen pop-up on first load turns a Discover tap into an immediate back-swipe.
5. Serve over HTTPS and keep the page indexable. Secure delivery is a page-experience signal, and indexing is the eligibility condition. A noindexed page is not a candidate no matter how good the image is.
6. Avoid images that are logos or mostly text. Google's guidance says not to use your site logo as the image. A card in a feed competes with photographs.

The editorial side: headlines, freshness and depth

Google is unusually blunt about headlines in the Discover documentation. It asks you to "use page titles and headlines that capture the essence of the content," to "avoid clickbait and similar tactics to artificially inflate engagement," and to "avoid sensationalism tactics that manipulate appeal."

Read that as a description of the enforcement. A headline that over-promises earns the tap and loses the reader, and Google can see both halves of that.

• Write the headline the article can cash. If the piece explains how the new tax slab affects a ₹12 lakh salary, say that. "You won't believe what the new slab does to your salary" is the pattern Google names as clickbait.
• Publish while the interest is live, and update when it moves. Google's phrase is "timely for current interests." Refresh the same URL rather than republishing — that is also the rule we apply site-wide.
• Go deeper than the feed around you. Discover's other favoured traits are "tells a story well" and "provides unique insights." Both are about depth, sourcing and a point of view — the E-E-A-T signals Google's people-first guidance describes, and the reason a thin rewrite of someone else's story rarely gets in.

If you publish regularly and want headline, image and freshness discipline built into the workflow rather than bolted on afterwards, that is the editorial layer our content marketing services are designed to run. The technical checklist is a one-time fix; the editorial side is a habit.

How to read the Discover report in Search Console

The Discover performance report shows you three numbers — clicks, impressions and average CTR — for content that appeared in Discover. There are no queries, because there were none, and there is no average position, because a feed has no positions (Search Console Help).

Two definitions change how you read the numbers. An impression means the card "was scrolled into view," counted once per session — so impressions are people who actually saw your card. And "a click is not counted if the user shares the item, or performs any other action" — shares and saves are invisible to you.

You can group by page, country, date and Discover appearance (AMP, News Showcase, Web Stories, videos). The report is "visible only if your property has reached a minimum number of impressions in Discover," and Google does not publish that minimum. If you do not see the tab, you have not crossed it yet.

How it sits alongside the standard Search report is covered in our Google Search Console guide. Three things to do the first time you open the Discover tab:

• Sort pages by impressions, then look at CTR. High impressions and low CTR is a card problem — usually the image or the headline.
• Compare the date graph to your publishing calendar. Discover spikes almost always trace to a specific article on a specific day. That is your evidence for what the feed wants from you.
• Filter by country. An India-focused site sometimes finds a surprising share of Discover impressions from the Gulf or the US — the diaspora reading Indian news.

Google has also added a separate generative AI performance report for Discover, tracking impressions your links receive inside AI-generated features in the feed. Read it separately.

Does Preferred Sources help with Discover?

Preferred Sources helps with Top Stories, AI Overviews and AI Mode — not, as far as Google has documented, with Discover.

The feature launched on 12 August 2025 "for all users in the U.S. and India," letting readers pick outlets they want to see more of in Top Stories (Google's launch post). Google's publisher documentation, updated 20 August 2026, says it is now global for Top Stories in every language Google Search supports and also applies to AI Mode and AI Overviews. Only domain-level and subdomain-level sites are eligible, and publishers can add a "preferred source" button with a two-line script or a deep link to google.com/preferences/source.

Discover is not mentioned in that documentation, so we do not claim an effect on the feed. What Google did announce on 20 August 2026 is a separate Discover control: readers will be able to "tap any three-dot menu in your feed to tell Google the exact topics or links you want to see more (or less) of" (The Keyword). Readers can already block a source with "Don't show content from [Source]" — so a misleading card does not just lose one tap; it can lose that reader permanently.

If you are a news or topical publisher, add the Preferred Sources button because it is free and helps in Top Stories and AI surfaces. Just do not count it as a Discover tactic.

Discover traffic is volatile — treat it as upside, not a plan

Google says this in its own documentation: "Traffic from Discover is less predictable or dependable when compared to keyword-driven search visits." It lists the reasons as changing user interests, content types, and updates to Google Search (Search Central).

Search traffic is a stock: you rank, people keep typing the query, and the traffic arrives every month. Discover traffic is a flow: an article catches an interest, the feed pushes it for a few days, the interest moves on, and the traffic drops to zero. Both are real. Only one is plannable.

• Forecast on Search, report Discover separately. Never blend the two in a monthly number — a Discover spike in March makes April look like a collapse.
• Do not build a team around Discover. A payroll that depends on feed spikes depends on Google's mood. Build around evergreen search demand and let Discover be the bonus.
• Treat a drop as information, not a crisis. Check whether interests moved, whether your content drifted toward clickbait, or whether a Search update landed — Google's three reasons, in that order.
• Reinvest the spike. The readers a hit brings are new. Give them a reason to return that does not depend on the feed — a newsletter, a link to your best evergreen page.

Where to go from here

Fix the six technical items first — the image size and the max-image-preview:large tag take an afternoon. Audit your last twenty headlines against Google's clickbait wording. Then open the Discover tab in Search Console, if you have it, and let the page-by-page data show which content shape the feed already likes from you.

If you do not have the tab yet, the work is identical to what earns rankings: our guides on how to rank on Google and content marketing strategy cover that foundation. Discover follows from it; it does not replace it.