Three of our 52 sites started appearing in Google Discover in March 2026. One of them received more traffic from Discover in a single week than it had received from organic search in the previous month. The content was not new — it was an article published six weeks earlier that suddenly matched enough interest signals to get surfaced.
That is how Discover works. It does not care about your keyword targeting. It does not care about your search intent optimization. It cares about one thing: whether its algorithm predicts that a specific human being will want to read your specific article right now.
And it reaches more than 800 million devices worldwide.
What Google Discover Actually Is
Google Discover is the content feed that appears on the Google app home screen, the Chrome new tab page on mobile, and google.com on mobile browsers. It surfaces articles, blog posts, videos, and web stories to users based on their browsing history, search history, location, and inferred interests.
Users do not search for anything. Content appears in their feed automatically. This makes Discover fundamentally different from Google Search — it is a push channel, not a pull channel. Content finds the reader, not the other way around.
Google has disclosed that Discover reaches over 800 million monthly active users globally. For publishers who crack the algorithm, Discover can deliver traffic volumes that rival or exceed organic search — with zero keyword research required.
Why Most Sites Never Appear in Discover
Discover has eligibility requirements that filter out the majority of the web:
Content policies. Your site must comply with Google's content policies and not have any manual actions. This eliminates sites with spam, thin content, or policy violations.
Indexing. Your content must be indexed by Google. Pages blocked by robots.txt, noindexed, or not yet crawled are ineligible.
Content quality. Google applies E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) signals aggressively in Discover. Sites without clear author attribution, about pages, or topical authority rarely appear.
Image requirements. This is where most sites fail silently. Discover strongly prefers content with large, high-quality images. Google's documentation specifies that images should be at least 1,200 pixels wide and enabled via the max-image-preview:large meta robots tag. Without this tag, your content is effectively invisible to Discover.
In our network, adding the meta robots tag was the single change that unlocked Discover eligibility for multiple sites that had never appeared before:
<meta name="robots" content="max-image-preview:large">
The Interest Graph: How Discover Ranks Content
Discover does not use PageRank or backlink signals the way Search does. It uses an interest graph — a model of each user's interests built from their activity across Google products.
The algorithm evaluates two things:
Topic relevance. Does the content match a topic the user has demonstrated interest in? Google tracks this through search history, YouTube watching patterns, location data, and interactions with previous Discover content.
Content quality signals. Is the content authoritative, well-produced, and engaging? Discover measures engagement through click-through rate and dwell time. Content that earns clicks but high bounce rates gets suppressed. Content with strong dwell time gets promoted.
This creates a feedback loop. When your article first appears in Discover, it is shown to a small test audience. If that audience engages (clicks and reads), it gets pushed to a larger audience. If they bounce, it gets killed. The first 2-4 hours after surfacing are critical.
Optimizing for Discover: What Actually Works
Emotional Headlines (Not Clickbait)
Discover rewards headlines that generate emotional engagement without being misleading. "How I Cut My Hosting Costs by 90%" works. "You Won't Believe How Cheap Hosting Can Be" does not.
The distinction is specificity. Discover's clickbait classifier penalizes vague, sensationalized headlines. It rewards headlines that make a specific, verifiable claim and deliver on it.
In our testing, headlines with numbers, first-person framing, and specific outcomes outperformed generic headlines by 2-3x in Discover click-through rates. The best-performing headline formula was: "[First-person action] + [specific result] + [timeframe or constraint]."
Topic Authority Over Keyword Targeting
Discover does not match keywords. It matches topics. A site that has published 15 articles about condo investing will appear in Discover for condo-related content far more frequently than a site that published one article about condos among 200 articles about unrelated topics.
This is topical authority in its purest form. The more consistently you publish on a specific topic cluster, the more likely Discover is to recognize your site as an authority in that cluster and surface your content to interested users.
Our sites that focus on a single niche (condo costs, home building, financial independence) appear in Discover at significantly higher rates than sites that cover broad topics.
Freshness Signals
Discover has a strong recency bias for news and trending topics, but evergreen content can also appear if it matches a user's interest at the right moment. We have seen articles published months ago suddenly appear in Discover feeds after a related news event or seasonal trigger.
To maximize freshness signals, update your sitemap's <lastmod> dates when you make meaningful content updates. Republish updated articles with a note at the top indicating the update. Discover treats meaningfully updated content as fresh.
Web Stories for Premium Placement
Google Web Stories get premium carousel placement in Discover — a dedicated, visually prominent section above the standard feed. Web Stories are essentially AMP-powered, full-screen, swipeable visual stories similar to Instagram Stories but hosted on your own domain.
Creating Web Stories with tools like makestories.io or the WordPress Web Stories plugin gives your content access to this premium Discover placement. In our testing, Web Stories received 5-10x more Discover impressions than standard articles on the same topics.
Image Optimization
Every article targeting Discover needs a hero image that is at least 1,200 pixels wide. Use WebP or AVIF format for fast loading. Ensure the image is relevant to the content — Discover's algorithm evaluates image-content relevance.
Add descriptive alt text and use the max-image-preview:large meta tag. Without this tag, Google may not show your image in Discover, which dramatically reduces click-through rates.
Monitoring Discover Performance
Google Search Console provides a dedicated Discover performance report. Navigate to Performance > Discover to see impressions, clicks, and click-through rates for your Discover appearances.
Key metrics to track:
- Impressions: How often your content appeared in Discover feeds. High impressions indicate strong topic matching.
- CTR: Discover CTR benchmarks are typically 4-12%. Below 4% suggests your headlines or images need improvement. Above 12% indicates strong engagement signals.
- Content type breakdown: Identify which articles trigger Discover appearances. Look for patterns in topic, format, headline structure, and publish timing.
Discover traffic is inherently variable. You may see spikes on some weeks and nothing on others. The goal is to increase the frequency and amplitude of spikes by consistently publishing content that matches the interest graphs of your target audience.
The Implementation Checklist
Here is what I did across all 52 sites to optimize for Discover:
- Added
<meta name="robots" content="max-image-preview:large">to every page template - Ensured every blog post has a hero image at least 1,200px wide
- Rewrote headlines to use the specific-outcome formula
- Added author schema with credentials to every article
- Verified topic clustering — each site focuses on 2-3 related topic clusters maximum
- Set up Discover reporting in Search Console for every property
- Created Web Stories for the top 5 performing articles on each site
Total implementation time: approximately 8 hours across the network. For a single site, 1-2 hours covers everything except Web Story creation.
Discover is not a channel you can force. You cannot buy your way in, and you cannot game the algorithm with tricks. But you can make your content eligible, optimize the signals that matter, and then let the interest graph do its work across 800 million devices.
For the complete strategy behind building sites that earn passive discovery traffic, check out The $100 Dollar Network.