Start a project

Google Discover: how a news site gets in and what counts

Google Discover: how a news site gets in and what counts

A news site gets into Google Discover automatically, with no application and no special code. Its pages just need to be indexed by Google and to follow Discover’s content policies. Whether a given article shows up depends on the user’s interests, the image, the title and the overall quality of the site. Everything below is based on Google’s own documentation (Google Search Central).

We have built news sites such as Ναυτεμπορική, ΕφΣυν, ΔΕΔΟΜΕΝΟ, newsbreak and δημοκρατία. Here you will find what Google asks for and how to apply it in practice.

What Google Discover is

Discover is the feed of articles a user sees in the Google app and in Chrome on mobile, without having searched for anything. Google picks what to show based on each person’s interests.

For a publisher, the difference from search is big. In search, the reader arrives with a question. In Discover, the article has to win them over inside a feed, with a title and an image.

In Greece, «google discover» gets an average of 320 searches a month, with a median of 260 (Google Keyword Planner via DataForSEO, November 2026). On 16/11/2026 every result on the first page was in English.

How much traffic Discover brings to a news site

The volume varies a lot from site to site and from week to week. Google itself says Discover traffic is less predictable than search traffic and should be treated as supplemental.

In practice that means two things. One article can bring a sharp spike within minutes, so the server has to cope with it. And a drop in Discover does not necessarily point to a quality problem. Google notes that traffic also shifts with its own updates and with users’ interests.

Don’t build a site’s revenue plan on Discover. Build it on steady traffic from search, social and direct visits. For how AI Overviews are changing search traffic, see our guide to AI Overviews.

Technical requirements for Discover

Google says no special tags or structured data are needed for a page to be eligible. The basics still have to be right.

Indexing

The article has to be in Google’s index. Check in Search Console that new articles get indexed quickly and that your sitemap updates with every story.

Large images

According to Google, Discover images should be at least 1200 px wide, more than 300,000 pixels in total and in a 16:9 ratio. Large images must be enabled with the max-image-preview:large setting or by using AMP.

The setting goes into a robots meta tag in the page head. Yoast SEO and Rank Math add it by default, but it is worth checking in the source, especially on custom themes.

Declare the main image with og:image or with schema.org markup. Google asks you not to use the site logo as the article image and to avoid text-heavy images.

Speed and page experience

Google asks for an overall good page experience. On a news site that means a fast LCP on mobile and a layout that doesn’t jump when ads load. Ad slots need reserved space, otherwise CLS goes up.

Speed during spikes depends on hosting. When an article takes off in Discover, the site gets many readers at once. Without cache and a CDN, the server slows down at the moment that matters most. More in our guide to web hosting.

Do you still need AMP?

No. Google lists AMP as one of two ways to get large images, alongside max-image-preview:large. With the setting in place, your site can do without AMP. One template instead of two means less maintenance and fewer errors in ads and analytics.

Structured data

Discover doesn’t require structured data. Article schema with a correct headline, dates, author and image still helps Google understand the page, and it serves search as well.

What content Discover picks and what it penalises

Here Google’s documentation is clear. It asks for titles that capture the essence of the article, and for content that is timely, tells a story well or offers its own insight.

It also asks you to avoid two tactics:

  • Clickbait: a misleading or exaggerated title and image, or holding back the key information to get the click.
  • Sensationalism: titles that play on morbid curiosity, titillation or outrage.

In Greek news publishing this covers some familiar patterns. A title that hides the person or the event goes against the guidelines.

AvoidPrefer
«See what happened in Parliament today»«Parliament passes the bill on [topic]: what changes»
«Shock over the announcement nobody expected»«[Company] closes [number] stores: the announcement»
«This is the secret to cheaper electricity»«How to cut your electricity bill this winter»

Google also notes that Discover may show older content if it is useful for a user’s interests. Guides and features have a place there, as long as you keep them up to date.

Trust and transparency

On quality, Google points to its guidance on helpful, reliable, people-first content. For a publisher that means bylines linked to author bio pages, and clear publication and update dates. About and contact pages count too.

Content policies

Discover uses SafeSearch and filters out content that might be seen as shocking. According to Google, it may not recommend job applications, petitions, forms or satire without context. If there is a violation, the manual action appears in Search Console under Security and Manual actions.

Google News and Publisher Center in 2026

Discover and Google News are different products. For Top stories and the News tab in search, publishers are considered automatically, with no application. Google finds sites through its normal crawl and asks for quality content that follows the Google News policies.

You use Publisher Center for your organisation’s settings, user management and tools such as Reader Revenue Manager. Google News Showcase is a separate programme, with its own agreement between Google and the publisher.

A news sitemap helps Google find new articles quickly. Google asks that it only lists articles from the last two days.

How to measure Discover in Search Console and GA4

Search Console

Search Console has a separate Discover report under Performance. It shows impressions, clicks and CTR for the last 16 months, and it includes traffic from Chrome.

The report only appears once the site passes a minimum threshold of impressions. A new site may not see it straight away. Use it to find which articles take off and what they have in common: topic, publishing time, image, title length.

GA4

In GA4, Discover traffic has no source of its own. Depending on the app and the browser, it is recorded as google / organic or as direct. That is why GA4 numbers don’t match Search Console.

The practical fix: read Discover clicks in Search Console. In GA4, measure what readers do once they arrive: how much they read and if they go on to a second article. For publishers, we build dashboards that show the newsroom GA4 and Search Console side by side.

Frequently asked questions

Does a site need to apply to get into Google Discover?

No. Any page that is indexed and follows Discover’s policies is eligible. Being eligible doesn’t guarantee that it will appear, though.

What image size does Discover need?

At least 1200 px wide, more than 300,000 pixels and a 16:9 ratio, with max-image-preview:large enabled (Google Search Central).

Why did my Discover traffic drop suddenly?

Discover traffic rises and falls with user interests and with Google’s updates. First check Search Console for a manual action, and look for technical changes to images or robots meta. If everything is in order, give it time.

Do I need AMP for Discover?

No. The max-image-preview:large setting is enough for large images to show.

Do older articles appear in Discover?

Yes. Google also shows older content if it matches the user’s interests. Guides and analysis have a place in the feed.

If you run a news site and want a Discover check, from images and meta tags to speed under spikes, see MINOANDESIGN SEO services or send us your site.

Have a project in mind?

Tell us where you want to go. We design, build, host and grow websites that perform.