Feed and Discover optimization

TL;DR

Feed and Discover placement rewards publishers that combine strong editorial signals with the technical hygiene the surface expects. Ship max-image-preview:large in the robots meta, 1200-pixel-wide 16:9 hero images, honest headlines with sharp story angles, and clean Article schema with named authors. Discover cannot be submitted to. It can only be earned.

The surface, in one paragraph

Google Discover is the personalized content feed that appears on Chrome mobile's home tab, in the Google app, and on Android home screens. It is not a search product. There is no query. Google picks stories it thinks a person wants to read based on inferred interests from search history and interactions. Apple News does the same job for iOS users, gated behind Apple's editorial and publisher program. MSN, Flipboard, and SmartNews aggregate feeds through their own selection engines. Together, these feeds account for a growing share of publisher traffic, sometimes exceeding organic search on volatile days. FEO is the discipline of making a publication eligible for these feeds and structuring individual stories to earn placement once eligible.

Where this fits in Search Everywhere Optimization

FEO borrows heavily from adjacent surfaces.

SEO shares the substrate. Discover pulls from Google's index. If the site is not indexed cleanly, Discover cannot find it. Fundamental SEO hygiene is a prerequisite.

E-E-A-T gates Discover eligibility. Sites without visible author identity, editorial standards, and trust signals do not enter the Discover pool. E-E-A-T is upstream.

CWV filters slow pages. Discover downweights slow mobile pages. LCP and CLS discipline pays off in feed placement.

VDO overlaps for video stories. Discover surfaces YouTube and web-hosted video. Chapters and VideoObject schema help.

KGO strengthens entity association. When a story is about a named entity with a Knowledge Panel, Discover's entity model surfaces it to that entity's followers.

The five levers

1. max-image-preview:large in the robots meta

Discover requires this one meta tag on every article page: <meta name="robots" content="max-image-preview:large">. Without it, the article is capped to a small thumbnail and rarely earns Discover placement. This is the single cheapest FEO fix. Most WordPress installs miss it because SEO plugins default to different values.

2. Hero image discipline: 1200 wide, 16:9 or larger

Discover cards show a wide hero image. The image needs to be at least 1200 pixels wide, ideally near 16:9 or wider. OG image tag points to this file. The photo is honest to the story: no lifestyle-stock-photo-and-clickbait mismatch. The visual has room for the Discover overlay UI at the bottom. I test the image at Discover-card scale before publishing.

3. Headlines that promise a real story

Discover's algorithm reads the headline as intent. Titles that describe the actual story, with a specific angle, earn CTR. "New study finds X" beats "You won't believe what happens next." Titles that mislead cost the publisher long-term eligibility. Editorial guidelines: subject, verb, specific object, ideally a name or a number. Discover title works best at 40 to 60 characters.

4. Article schema with named author and datePublished / dateModified

Article, NewsArticle, or BlogPosting schema on every story. author points to a Person entity with sameAs and knowsAbout. publisher points to the Organization. datePublished and dateModified in ISO 8601. Discover privileges fresh content, and dateModified is how it detects updates.

5. Publisher signals and Google News Publisher Center registration

Google News Publisher Center registration (formerly Google News). Site verified. Publisher logo uploaded at correct dimensions. Categories chosen accurately. This step qualifies the site for Google News and improves Discover eligibility as a side effect. For Apple News, publisher application through Apple News Publisher account, Apple News Format if publishing at scale.

First 30 / 60 / 90 days

Days 1 to 30: audit and prerequisites

Discover eligibility audit. Does the site appear in GSC's Discover report. Historical impressions and CTR.

Technical audit. Is max-image-preview:large set on every article. Is the OG image tag pointing to a 1200+ wide image. Is Article schema clean. Is CWV passing on article templates.

Publisher signals. Google News Publisher Center account state. Apple News publisher status. About page and editorial standards page.

E-E-A-T posture. Named authors. Person schema. Corrections policy. Editorial guidelines page.

Content and headline audit. Are headlines specific and honest. Are hero images honest to the story. Any clickbait patterns.

Deliverable at day 30: eligibility scorecard, technical fix queue, publisher registration status, editorial standards remediation list.

Days 31 to 60: fix and register

max-image-preview:large deployed site-wide. Verified in raw HTML on every article template.

Hero image workflow updated. Photo editors and CMS templates enforce 1200+ wide 16:9 hero images. OG image tags autopopulate.

Article schema shipped with author, publisher, datePublished, dateModified. Named authors linked via Person schema.

Google News Publisher Center registration and site claim. Publisher logo uploaded at 1000x1000 (square) and other required dimensions.

Apple News publisher account created if the volume and topic support it. Sitemap or RSS submitted per Apple's onboarding.

Editorial standards and corrections policy pages published.

Deliverable at day 60: full technical compliance, live publisher registrations, editorial pages published, ready-to-earn baseline.

Days 61 to 90: measure and iterate

GSC Discover report review. Which stories earned impressions. Which headline patterns won CTR. Which topics converted to sessions.

Editorial iteration. Headline formulas that work extended. Hero image treatment refined. Story angles that surfaced repeated.

Apple News stats review if the publisher is on the platform.

MSN and Flipboard eligibility. If the site is strong on quality and topical, apply.

Freshness discipline. Every legacy article that gets updated bumps dateModified with a real update note.

Deliverable at day 90: measured Discover lift, working headline and hero image formula, live publisher registrations, editorial calendar tuned for feed placement.

Tools I use

• Google Search Console. Discover performance report, Google News tab.
• Google News Publisher Center. Publisher registration, logo management, category configuration.
• Apple News Publisher account. Publisher onboarding, Apple News Format tooling.
• Rich Results Test. Article and NewsArticle schema validation.
• PageSpeed Insights. CWV on article templates.
• Yoast, RankMath, or headless SEO plugin. max-image-preview:large enforcement.
• ScreamingFrog SEO Spider. Site-wide meta robots audit.
• Flipboard publisher tools. Feed submission and category placement.
• Ahrefs Discover keyword database. Which topics earn Discover impressions competitively.
• NewsWhip. Real-time trending story detection for topical timing.

What kills the program

Missing max-image-preview:large. Without it, Discover eligibility drops sharply. Every article template must ship it.

Clickbait headlines. Sites lose Discover placement fast when the algorithm detects mismatch between headline and content.

Stock-photo heroes. Discover downweights obvious stock imagery. Original photography or commissioned illustrations perform better.

Slow mobile pages. Failing CWV cuts Discover impressions. LCP over 4 seconds is a de facto exclusion.

Anonymous articles. No named author, no Person schema, no editorial standards page. Discover downweights.

Chasing Discover as the main channel. Traffic vanishes for days at a time. Publishers who build P&L around Discover blow up.

Stale content. dateModified without real updates is detectable. Publishing thin "refreshes" loses eligibility.

Ignoring Apple News. On lifestyle, health, and news topics, the Apple News audience is real. Publishers who skip Apple leave meaningful reach on the table.

KPIs that matter

• Discover impressions in GSC. Monthly and weekly trends.
• Discover CTR. Average and per-article, indicating headline/image quality.
• Discover sessions in GA4. Referrer google.com with source=discover parameter.
• Apple News impressions and read count. Apple News Publisher analytics.
• Publisher trust score. Composite of author coverage, editorial standards presence, corrections logged.
• Hero image compliance rate. Percent of new stories meeting the 1200-wide 16:9 standard.
• Discover topic mix. Which topical clusters produced the most impressions.

FAQ

What is Google Discover?

Google Discover is the personalized feed of articles Google shows on the Chrome mobile home page and the Google app. It is entirely algorithmic. There is no query. Discover pushes content to users based on interests inferred from search history.

How do I get onto Google Discover?

You cannot submit. Sites become eligible when they meet Google's content quality bar and publisher signals. Once eligible, the next step is optimizing image ratios, headlines, and freshness to earn impressions.

What image ratio does Discover want?

1200 pixels wide minimum, aspect ratio near 16:9 or larger. I set max-image-preview:large in the robots meta tag and use OG image tags at these dimensions.

Is Apple News worth the effort?

For publishers, yes. Apple News reaches over 125 million users. The Apple News Format publisher program requires editorial standards and platform submission. For non-publishers, the ROI is thinner.

How volatile is Discover traffic?

Very. A single article can drive 10x the site's normal traffic for a day then vanish. I plan editorially and financially assuming Discover is bonus, not baseline.

Does clickbait work in Discover?

Short term, yes. Long term, no. Google explicitly downweights clickbait titles and misleading images. Sites that go clickbait for a quarter lose Discover eligibility.

Related reading

• E-E-A-T playbook: the trust prerequisites for Discover eligibility.
• Core Web Vitals: mobile speed gates Discover placement.
• Video search optimization (VDO): Discover surfaces video stories too.
• Knowledge graph optimization (KGO): entity signals help topical eligibility.
• Visual search optimization (VxSO): hero image discipline overlaps.