Google Discover SEO: How to Earn Traffic From Discover in 2026

Maybe your traffic from Google Discover jumps one week and vanishes the next. Or maybe you have never seen any at all. You are not alone. In my experience, Google Discover SEO is less about ranking for keywords. It is more about publishing content that Google’s systems think a person will want to read before they even search.

The short answer: you cannot submit pages to Discover. Indexed pages that follow Google’s content policies can show up on their own. What you can control is the quality and timing of what you publish. You also control the image, the title, and how good the page feels to use.

Below I cover what Google’s own “Get on Discover” guidance asks for, how to read your Discover data in Search Console, and why the traffic swings so much. I also cover what to do while the first Discover core update is still rolling out.

What Google Discover Is and How It Picks Content

Discover is the personal feed of content cards in the Google app and on the Google homepage in mobile browsers. It does not answer a search. Instead, it suggests articles, videos and other content based on what Google knows about a person’s interests.

Those interests come from signals like search activity and the topics and sources a person interacts with. People can also shape their own feed. Since September 2025, signed-in users can follow publishers and creators right inside Discover. They can also open a separate space for that source’s content.

Google has also started mixing in posts from platforms like X, Instagram and YouTube Shorts. That matters if you publish articles. Your articles now compete with social posts and videos for the same space in the feed.

Discover Is Only Loosely Tied to Search Rankings

This is the part I see most people get wrong. Ranking on page one for a keyword does not put you in Discover. A page with no rankings at all can still be shown there. There is no search to rank for, so keyword targeting does very little.

Still, Discover is part of Google Search. Google’s documentation says Discover uses many of the same signals and systems as Search to decide what is helpful, people-first content. So the quality signals overlap, but the trigger is different. Discover works from interest, not from a search.

Here is the simple way I think about it. Search answers “what did this person ask?” Discover answers “what would this person want to see right now?”

Is Your Site Eligible for Google Discover?

Eligibility is simpler than most guides make it sound. Google says content shows in Discover on its own if it is indexed and follows Discover’s content policies. You don’t need special tags, structured data or a sign-up.

Google is also clear that meeting the rules does not guarantee anything. Being eligible gets you into the pool. It does not earn you a place in anyone’s feed.

Before I work on anything else, I check three basics:

• Indexing: your articles are indexed (stored by Google so they can be shown). Use the URL Inspection tool in Search Console on a few recent posts.
• No accidental blocks: pages don’t carry a stray noindex, and your robots rules don’t limit image previews.
• No manual actions: Google reports Discover policy problems in Search Console under Security & Manual Actions. Check there if traffic stops completely.

Google Discover SEO: What Google’s Guidance Actually Asks For

Google’s Get on Discover documentation is short. I take every point in it seriously. Here is each one, and what it means in practice.

Use Large, High-Quality Images

Discover is a visual feed, so the image does a lot of the work. Google recommends images that are:

• At least 1,200 pixels wide.
• High resolution, with more than 300,000 pixels in total. A 1,280 x 720 image, for example, is well over that.
• In a 16:9 shape where possible. Google crops images on its own. If you crop by hand, keep it wide, not tall, and keep the important detail in the frame.

Size alone is not enough. You also need to let Google show the image large. Google asks you to allow large previews, either with the max-image-preview:large robots setting or by using AMP. Without it, your card can show a small thumbnail. That tends to get far less attention in the feed.

<meta name="robots" content="max-image-preview:large">

Not sure how this rule works next to noindex, nosnippet and the rest? Our guide to robots meta tags and the X-Robots-Tag header explains how each one works and how they combine.

Next, tell Google which image stands for the page. Use the og:image meta tag or the image property in your schema.org markup (code that describes the page to search engines). Point it at a real photo or drawing of the story.

Google warns against generic images like your site logo, and images with lots of text on them. A recipe article should lead with the finished dish. It should not lead with a branded banner that has the title written across it.

Write Titles That Describe the Content, Not Tease It

Google changed its title advice in early February 2026, at the same time as the Discover core update news. It now says to use titles and headlines that capture the core of the content. It also says to avoid clickbait and similar tricks that push up engagement in a fake way.

It also asks you to avoid sensational tricks that play on people’s feelings. And it asks you not to mislead with exaggerated details in titles, snippets and images. Hiding the key fact so people have to click is exactly what it means. Honestly, I think this is the change that matters most for many news sites.

Here is the difference in practice, using a made-up local news story:

• Clickbait: “You won’t believe what the council just did to parking.”
• Descriptive: “Council raises city-centre parking charges by 20% from April.”

The second one tells readers exactly what they will get. It still gets clicks from people who care, and nobody feels tricked.

Publish Timely Content, or Content With a Fresh Angle

Google asks for content that is timely for what people care about now, tells a story well, or gives a unique view. Discover rewards things people are interested in right now. That could be news, a new product, a seasonal topic, or an issue many people are talking about.

Timely does not only mean news, though. Discover can show older content if it is still useful and matches someone’s interests. A solid, evergreen guide (one that stays useful over time) can keep showing for months when interest in its topic rises again.

What rarely works is a thin rewrite of a story everyone else has already published. If your piece adds nothing new, Discover has little reason to pick yours. When I plan a Discover piece, I always ask first: what does this add that other versions don’t?

Show Expertise and Trustworthiness (E-E-A-T)

The Discover documentation points site owners to Google’s guidance on helpful, reliable, people-first content. That is the same guidance where Google talks about experience, expertise, authoritativeness and trust (E-E-A-T).

For Discover, it comes down to a few things people can see:

• Clear bylines that link to author pages. These show who wrote the piece and why they know the topic.
• Visible publish dates, plus update dates when a piece has been changed.
• An about page and contact details that make it obvious who runs the site.
• Original reporting, hands-on testing or analysis, not just summaries of other people’s work.
• Topic focus. A site that covers one subject in depth is easier to trust on that subject than one that covers everything.

Provide a Good Page Experience

The same February change added advice to give people a great page experience overall, with a link to Google’s page experience guidance. In practice, that means fast mobile pages, no annoying pop-ups covering the article, and ads that don’t bury the content.

Remember that Discover is mostly used on phones. Someone who taps a card and lands on a cluttered page that jumps around as it loads is unlikely to come back. That reflects badly on what you offer. I always open a few articles on my own phone before I worry about anything else here.

Discover Content Policies and What Is Excluded

Discover follows the Google News content policies. These cover areas like dangerous content, deceptive practices, harassment, hateful content, manipulated media, medical content, sexually explicit material, violence and gore, and vulgar language.

The transparency policy is the one that catches normal publishers out. Readers should be able to see who wrote the content and who publishes it. That means clear dates, bylines and ways to contact the publisher.

Google also says Discover generally does not show some types of content, including:

• Job postings
• Petitions
• Forms
• Code repositories
• Satire without context that explains it is satire

Discover also uses SafeSearch. It filters out content that may be shocking or unexpected in a general feed, even if that content is allowed in normal search results.

The Follow Feature, RSS Feeds and Web Stories

Older guides still tell you to tune your RSS feed for Discover’s “Follow” button. That advice is out of date. In November 2025 Google noted that the older Follow feature was no longer shown in Discover. That feature used RSS and Atom feeds and first started in Chrome. Google also removed the feed advice from its Discover documentation.

Following now happens inside Discover itself. Signed-in users can follow publishers and creators by name. I still keep a working RSS feed, because other readers and tools use it. But there is no special feed setup for Discover.

Web Stories are a similar case. Google’s Web Stories documentation says a Story can still show in Discover as a single card that people tap to open. The special Web Stories carousel that many publishers used to build for has mostly faded away. Google has not said Stories are needed for Discover.

If you already make Web Stories and people engage with them, there is no reason to stop. If you don’t, you are not missing a required Discover format.

How Discover Traffic Shows in Search Console

Discover clicks don’t show in the main Search results performance report in Search Console. They have their own report, under Performance, called Discover.

A few things to know about it:

• It only appears after a minimum. If your site hasn’t had a minimum number of Discover impressions, the report won’t show at all. A missing report usually just means no real Discover traffic yet.
• The numbers are limited. You see clicks, impressions and click-through rate, broken down by page, country, appearance and date. There are no queries, because nobody searched.
• It covers 16 months. That is enough to compare seasons and see whether a drop is new or part of a pattern.
• It includes Chrome. Discover impressions in Chrome are counted in the report.

Google Analytics 4 has no separate Discover channel by default. So Search Console is the most reliable place to see how Discover is doing. For any page that gets a lot of Discover traffic, compare its Discover clicks with its Search clicks over the same dates. This is the first report I open when someone asks me why their traffic spiked.

Why Discover Traffic Is so Volatile

Discover traffic is spiky by nature. One article can bring more visits in two days than the rest of the site gets in a month. Then it can fall back to almost nothing. Google gives three reasons for this.

1. People’s interests change. Discover adjusts as interests shift, partly based on search activity. So the audience for your topic grows and shrinks.
2. The content mix changes. Discover often shows content types like sports, health, entertainment and lifestyle from across the web. The balance between them shifts over time.
3. Search updates. Discover is part of Search, so ranking and system updates can change your Discover traffic. Google’s guidance says the traffic can change no matter how good your content is or how often you publish.

Google’s advice is to treat Discover as an extra on top of search traffic, not something to build a business on. I agree with that fully. If a big share of your visits comes from Discover, that is a risk to manage, not just a win to celebrate.

The February 2026 Discover Core Update Is Still Rolling Out

On 5 February 2026 Google started its first core update made just for Discover. It is rolling out first to English-language users in the United States. Other countries and languages will follow in the coming months. Google said the rollout could take up to two weeks.

Google said the update aims to show more local content from sites based in the user’s country. It also aims to cut sensational content and clickbait, and to show more in-depth, original and timely content from sites with real expertise in a topic. Our overview of the first Discover core update covers the news in more detail.

At the time of writing, Google has not confirmed the rollout is complete. So any changes you see in your Discover report are not final. Don’t rewrite large parts of your site based on a few days of movement.

Wait until Google confirms it is done. Then compare a full period after the update with a similar period before it. The method in our December 2025 core update checklist works just as well for Discover. Compare the same kind of dates, look at which pages and topics moved, and judge your content quality honestly.

A Simple Discover Workflow for Your Next Article

This is the routine I’d build into publishing, so every article has a fair chance of being picked up.

1. Check the basics once. Add max-image-preview:large across the site, check your articles are indexed, and make sure there are no manual actions in Search Console.
2. Pick a topic people care about now. Choose something your readers care about today, or a fresh angle on an ongoing story, inside the subject your site is known for.
3. Write a clear title. Say what the article gives the reader. Remove teaser wording, hidden details and exaggeration.
4. Choose a strong main image. Use a relevant photo at least 1,200 pixels wide, ideally 16:9, with no heavy text on it and no logo-only images.
5. Declare the image. Set it as the og:image and in your Article structured data, so Google knows which image stands for the page.
6. Show who is behind it. Add a byline linked to an author page, a visible date, and original insight that other versions of the story don’t have.
7. Review results in Search Console. Check the Discover report every week. Note which topics and image styles get clicks, and use that to guide future articles.

Common Discover Mistakes to Avoid

• Chasing Discover with keyword stuffing. There is no search to match. Packing titles with keywords only makes them worse to read.
• Using the logo as the featured image. It is the default on many CMS templates, and it is exactly what Google says to avoid. I see this one all the time.
• Limiting image previews. Some SEO plugins or old templates set max-image-preview to standard or none. Check your page source.
• Publishing one story many times. Several near-identical articles on the same news split attention and add little value.
• Reacting to one bad week. A drop after a spike is normal. Look at weeks and months, not days.

Conclusion

Good Google Discover SEO comes down to doing the basics well, again and again. That means indexed pages, large images with max-image-preview:large, honest titles, real expertise and a clean mobile page. There is no submission form and no keyword trick.

I treat Discover traffic as a bonus on top of steady search traffic, and I expect it to rise and fall. Your next step: open the Discover report in Search Console and check your image settings. Then follow our coverage of the Discover core update so you know when it is safe to judge the results.