Google Discover Optimization: What It Is and What Actually Moves Traffic

SearchBlueprint · 20 September 2026 · 12 min read · scored 92/100

A smartphone displaying Google Search trends on a table at night.

Google discover optimization is the practice of shaping your content to earn visibility in Google's Discover feed. It's not a product you buy, it's work you do on your own pages. Effective optimization comes down to five levers. Use images at least 1,200px wide with the max-image-preview:large tag. Write curiosity-driven (but honest) headlines. Keep Core Web Vitals strong. Build topical authority. Publish when your audience is active. Discover requires no special markup. Eligibility is automatic once you're indexed and policy-compliant. But visibility is earned.

That gap between "eligible" and "visible" is where most publishers lose. Raptive sees around one billion Discover clicks per year across its network. It reports that for news sites, 76% of all Google traffic comes from Discover. Yet many publishers still treat it as an afterthought. This guide covers how the feed selects content, the exact specs Google publishes, the fixes with documented traffic gains, and how to measure results.

Key Takeaways

Discover eligibility is automatic for indexed, policy-compliant content. No special tags are required. But appearing in the feed is never guaranteed.
Images drive clicks. Use 16:9 images at least 1,200px wide with over 300,000 total pixels, and enable max-image-preview:large. One Google case study reported a 333% boost in clicks from that tag alone.
Discover behaves more like a social feed than search. Headlines that spark curiosity outperform dry, factual ones. Misleading previews violate policy, though.
Site authority matters more than in search. Raptive describes a "cliff" where sites below a certain authority threshold don't appear at all.
Measure in Search Console's Discover performance report, not Google Analytics. Analytics lumps Discover visits into direct traffic.

How Google Discover Optimization Actually Works

Google Discover is a personalized feed of articles and videos. It appears in the Google app, on Android home screens, and in mobile Chrome. It has been around since 2018. Unlike search, no one types a query. Google predicts interest from a user's Web & App Activity, then serves matching content.

According to Google's Discover documentation, content is automatically eligible if it's indexed and meets Discover's content policies. No structured data or special tags are required for eligibility itself. Discover uses many of the same signals as search to judge whether content is helpful and people-first.

There is no keyword intent to match. Users scroll one-handed and decide in a split second whether to tap. That makes the image-plus-headline pairing the single biggest on-page factor, much like a Facebook feed. Newsifier, a news CMS provider, says that in their experience 90% of news publishers already get more traffic from Discover than from Google Search. Most just don't realize it, because analytics tools misattribute the visits.

The authority cliff

Raptive's analysis found that domain authority and branded search volume correlate with both search and Discover, but differently. Search rewards authority roughly linearly. Discover has a cliff: sites below a certain authority level barely appear at all. If your site is new or unknown, build brand signals first. People searching your brand name, quality backlinks, and returning readers all tell Google your site belongs in the feed.

Which verticals win in Discover

Discover traffic is not distributed evenly. Raptive's network data shows dramatic differences by niche:

Vertical	Share of Google traffic from Discover	What it means for you	Source
News	76%	Discover is the primary Google channel; optimize every post for the feed	Raptive network data
Tech, gaming, sports, travel	~50%	Discover and search deserve equal investment	Raptive network data
Food	2%	Treat Discover as a bonus; search and recipe formats still dominate	Raptive network data

If you publish in a low-Discover vertical, don't abandon the feed, adapt formats. Raptive reports that roundup posts brought one recipe site nine times as much Discover traffic as a standard recipe. The feed rewards browsable, curiosity-driven formats even in verticals it usually ignores.

Smartphone displaying Google search page on a vibrant yellow background.

Image and Headline Specs That Google Publishes

This is the most concrete part of Discover optimization, because Google documents the requirements directly.

Image requirements

Google's guidance is specific:

At least 1,200 pixels wide. Smaller images may still appear, but as tiny thumbnails that get far fewer taps.
More than 300,000 total pixels. A 1280×720 image at 16:9 has 921,600 pixels and clears the bar comfortably.
16:9 aspect ratio. Google auto-crops for the feed. If you crop manually, keep key details inside the horizontal frame.
Enable large previews via the max-image-preview:large robots setting or AMP.
Specify your preferred image with og:image or schema.org markup. Avoid logos, generic stock photos, and text-heavy graphics. Google explicitly warns against all three.

The payoff can be dramatic. A Google case study cited by Newsifier reported that using the max-image-preview tag produced as much as a 333% boost in clicks. Raptive separately reports a site that grew Discover traffic 400% after changing a single meta tag.

Beyond specs, image choice matters. Newsifier compared two articles from the same site. One with an edited, emotionally expressive image earned a 21% click-through rate and 42,000 Discover visitors. A similar article with a dull stock photo managed 3%. Images with human faces, especially faces showing emotion or looking at the camera, consistently outperform object shots.

Headlines: curiosity without deception

Discover headlines work under different rules than search titles. There's no query to match, so keyword placement matters less than tap appeal. Raptive notes that title length in Discover runs longer than in search. That gives you room to be more descriptive and more creative. One food site in their network went from under 5,000 Discover clicks a month to over 150,000 after rewriting titles specifically for the feed.

The line to respect: Google's Discover content policies prohibit misleading preview content, promising a story the article doesn't deliver. Since the Helpful Content Update in August 2023, Google has also penalized sites whose headlines overpromise relative to the article. Research from Discover specialists also shows that content with explicit negative sentiment performs poorly in the feed. So write headlines that create a genuine curiosity gap, carry some emotion, and are fully paid off by the article. You can consider setting a distinct og:title for Discover while keeping a search-optimized title tag.

Technical Foundations: Core Web Vitals and Page Experience

Close-up of the Google homepage on a screen showing search options.

Discover is a mobile-only surface, so mobile experience is non-negotiable. Sites that load slowly or shift around while rendering struggle in the feed regardless of content quality. Raptive documents one site where fixing Core Web Vitals issues alone drove a 200% increase in Discover traffic.

Priorities, in order:

Time to first byte (TTFB). Discover specialists including Lily Ray single this out as especially important. A slow server response delays everything downstream.
Largest Contentful Paint. Your hero image is usually the LCP element, the same image doing Discover duty. Serve it in a modern format, properly sized, with high fetch priority.
Layout stability. Ads and late-loading embeds that shove content around kill engagement. Engagement signals feed back into Discover distribution.
Robots and social meta tags. Confirm max-image-preview:large is present sitewide, og:image points to a compliant image on every article, and nothing blocks indexing.

Because eligibility rides on indexing, verify each post is actually indexed quickly. Discover's window is short, more on that below. An article stuck in the crawl queue for two days misses its moment entirely.

Documented fixes and their reported results

Change made	Reported result	Why it works	Reported by
Enabled max-image-preview:large	Up to 333% more clicks in one Google case study	Large image cards get dramatically more taps than thumbnails	Google case study via Newsifier
Changed a single meta tag	400% growth in Discover traffic	Unlocked large previews the site was suppressing	Raptive
Fixed Core Web Vitals issues	200% increase in Discover traffic	Page experience is a distribution signal in Discover	Raptive
Rewrote titles for Discover (longer, curiosity-led)	From under 5K to over 150K monthly clicks	Feed users tap on intrigue, not keyword-matched titles	Raptive
Shifted publishing toward high-engagement days	Posts on Fridays were 6% of output but 22% of impressions	Discover rewards content that lands when demand peaks	Raptive

Each of these is a reported result from a specific site, not a guarantee, but together they show where the leverage sits: previews first, speed second, timing third.

Timing, Freshness, and Publishing Cadence

Discover strongly favors fresh content. Raptive observes that posts typically surface in the feed within 12 hours of publishing, if they surface at all. That short window has three practical consequences:

Publish when demand is high. One site discovered only 6% of its posts went out on Fridays, yet those posts earned 22% of its Discover impressions. Check your own Search Console data for day-of-week patterns and shift your calendar accordingly.
Refresh evergreen winners. Updating an older post with genuinely new information and a current timestamp can restore Discover eligibility, provided the update is real, not cosmetic.
Make indexing fast. Sitemaps, clean internal linking, and a valid RSS or Atom feed all help Google find new posts inside that 12-hour window. An RSS feed plus Organization and WebSite structured data also supports Discover's Follow feature. That feature lets Chrome and Android users subscribe to your site directly.

E-E-A-T and Topical Authority in the Feed

Discover leans on the same helpfulness and trust systems as search, per Google Search Central. But the authority cliff makes E-E-A-T signals even more decisive here. Practical moves:

Double down where you already win. Pull your top 25 Discover URLs from Search Console and look for topic clusters. Discover recommends content based on recognized authority; feeding it more of what it already trusts you on compounds results.
Build author entities. Real bylines, author pages, and consistent expertise across a beat help Google connect your content to credible people.
Cultivate returning readers. Newsletters, push notifications, and direct visits signal that users value your site. Discover favors sites users already engage with.
Stay family-friendly. Discover filters shocking, NSFW, and borderline content aggressively. Mixing edgy and mainstream content on the same subdomain can suppress everything.

The same clarity that earns trust in Discover also earns citations in AI answers. Writing self-contained, quotable passages is the core skill behind answer engine optimization, and it pays off in both channels at once.

Measuring Google Discover Optimization in Search Console

Do not trust Google Analytics here. Discover visits are counted as direct traffic. That's why Newsifier's team says 90% of news publishers underestimate, or entirely miss, their Discover volume. A "mysterious" spike in direct traffic is often Discover.

The reliable source is Search Console's Discover performance report. It shows impressions, clicks, and CTR for all content that appeared in Discover over the last 16 months. Data appears once you cross a minimum impressions threshold. The report includes Chrome traffic and covers every platform where users see the feed.

Use it to answer three questions on a recurring cycle:

Which topics earn impressions? That's Google telling you where it sees your authority. Publish more there.
Which pages get impressions but poor CTR? Those are image and headline problems. Fix the preview, not the article.
Which days and formats overperform? Reallocate your calendar toward them.

Treat Discover traffic as complementary and volatile, Google's own documentation says exactly that. Interest shifts, content-type adjustments, and core updates all move the numbers independent of anything you did.

What's Changing in 2026

Discover is evolving fast, and the recent shifts matter for planning:

First-ever Discover core update. On February 5, 2026, Google announced a core update focused specifically on Discover. It aims to surface more locally relevant content, reduce clickbait, and favor trusted sites. That raises the cost of borderline headline tactics.
AI summaries in the feed. Google began testing AI summaries in Discover in July 2025; Marfeel data cited by Raptive puts them at 51% of the feed. Your content can now be summarized before anyone taps. Making your pages easy for AI systems to represent accurately, the discipline behind generative engine optimization, is becoming part of Discover strategy, not separate from it.
Volatility from spam updates. The August 2025 spam update reshuffled Discover heavily; Raptive spoke to one creator who fell from 10 million monthly Discover clicks to under one million. The same update shifted traffic toward smaller independent publishers. Network-wide, Raptive saw non-food Discover traffic down 4% year over year in 2025, but up 5% in the second half. That's a reminder the feed rewards adaptation.
First-person content rising. Per Newzdash data, X became the fourth most popular domain in Discover in November 2025, behind YouTube, Yahoo, and CNN, echoing Reddit's rise in search. Authentic, opinionated, experience-based content is gaining share.
Personalized intelligence. Google launched personalized intelligence in Gemini and AI Mode in January 2026. It will reshape feed personalization in ways nobody can fully predict yet.

The through-line: Discover is converging with AI-driven discovery surfaces. Pages built with clear structure and liftable answers, the kind an AI can quote cleanly, are exactly what answer-engine-ready page design produces. They hold up better as summaries take over more of the feed.

Mistakes That Keep Sites Out of the Feed

Treating Discover like search. Keyword-stuffed titles and query-matching intros are wasted effort in a feed with no queries.
Chasing Discover before earning authority. Below the authority cliff, tactical tweaks accomplish little. Build brand search volume first.
Publishing preview bait. Exaggerated headlines and misleading images can trigger manual actions, visible in Search Console under Security & Manual Actions.
Building a business on Discover. The traffic is real but unstable, witness the 10x swings after the August 2025 spam update. Google itself calls Discover complementary to search traffic. Plan accordingly.
Ignoring the preview entirely. The inverse mistake: great articles with logo thumbnails and flat titles. In Discover, the preview is the product.

FAQs

Do I need special tags or structured data to appear in Google Discover?

No. Google's documentation states that indexed content meeting Discover's content policies is automatically eligible, with no special tags required. However, the max-image-preview:large setting and an og:image tag strongly influence whether your content gets a large, clickable card. One Google case study tied the preview tag to a 333% boost in clicks.

How long does it take for an article to show up in Discover?

Fresh content typically surfaces within about 12 hours of publishing, according to Raptive's network observations. If a post doesn't appear in that window, it usually won't later. That's why fast indexing and publishing on high-engagement days matter so much.

Why doesn't Google Analytics show my Discover traffic?

Google Analytics attributes Discover visits to direct traffic, so the channel is invisible there. Use Search Console's Discover performance report instead. It shows impressions, clicks, and CTR for the last 16 months once your site crosses a minimum impressions threshold.

Are clickbait headlines allowed in Discover?

Curiosity-driven headlines perform well and are allowed; deceptive ones are not. Google's policies prohibit previews that promise a story the content doesn't deliver, and the February 2026 Discover core update explicitly targets clickbait. Write intriguing headlines the article fully pays off.

Is Discover worth optimizing for outside of news?

It depends on your vertical. Raptive data shows Discover supplies roughly 50% of Google traffic for tech, gaming, sports, and travel sites, but only 2% for food. Even in weak verticals, format changes help, roundups earned one recipe site nine times the Discover traffic of standard recipes.

Check Your Discover Readiness Today

Pull up Search Console, open the Discover report, and audit your last ten posts against three questions: is the featured image at least 1,200px wide, is max-image-preview:large enabled, and would you tap that headline in a feed? Those three checks catch most of what holds sites back from google discover optimization gains. If you'd rather see the full picture, page experience, meta tags, authority signals, and content structure scored in one place, talk to the team and review a sample SEO report to see exactly what an audit of your site would surface.

Resources / References

Sources

About the author

SearchBlueprint researched, wrote and scored this article with the content pipeline our customers run on their own sites. The score in the byline is what our content engine gave this page before it published. Nothing here is reviewed by hand, which is the point: it is the product's output, published where anyone can check it.