Google Discover is a content feed displayed on the Google app and Android devices, surfacing articles, videos, and other content based on users’ interests, search history, and behavioral signals — without any search query. Unlike traditional search (intent-driven, keyword-matched), Discover is discovery-driven: Google predicts what users might find interesting and shows it proactively.
For publishers, Discover can be a significant traffic source. It sends volume in spikes — articles that Discover picks up can drive tens or hundreds of thousands of views in 48–72 hours. But the mechanics are opaque compared to traditional search, and the traffic is less predictable. Understanding what signals influence Discover surfacing helps content teams optimize for it alongside traditional SEO.
How Google Discover Differs from Search
No keyword intent: Discover doesn’t match pages to queries. It matches content to user interest profiles built from their Google account activity — what they’ve searched, what apps they use, their location, their YouTube watch history.
Freshness matters more: Discover heavily favors recent content. Articles published in the last 24–72 hours are most likely to be surfaced. Evergreen content can still appear in Discover but competes harder against fresh content.
Engagement signals dominate: Click-through rate on the Discover card and whether users actually read the article (time spent, scroll depth, return visit) are primary signals. Discover optimizes for content users engage with, not just click on.
No position tracking: There’s no equivalent of rank tracking for Discover. Articles are surfaced to users who find them interesting — you can’t see position or reliably predict when something will be surfaced.
Volume is spiky: A post that hits Discover can generate 10x its normal traffic in a short window, then return to baseline. Discover traffic isn’t consistent in the way organic search traffic can be.
What Makes Content Discover-Eligible
Google Search Console’s Discover performance report shows whether your content is being surfaced. For content to be eligible:
The site must be indexed by Google. This is the baseline — non-indexed sites can’t appear in Discover.
Content must meet Google’s content policies. Misinformation, adult content, and policy-violating content are excluded.
The page must have a high-quality large image. Google’s documentation explicitly states that Discover features large, compelling images that are at least 1200px wide. Content without qualifying images is rarely surfaced prominently. The image should be enabled using the max-image-preview:large robots meta tag:
<meta name="robots" content="max-image-preview:large">
Core Web Vitals are a stated Discover signal. Google has confirmed that page experience signals (including CWV) factor into Discover ranking, similar to their role in search.
Content Characteristics That Perform in Discover
Fresh, newsworthy topics: Content about current events, trending topics, and timely developments within an interest area performs well. Discover users receive content about topics they care about, but the freshness dimension means recent angles on those topics dominate.
Compelling, non-clickbait headlines: Discover card headlines must be engaging enough to earn a click, but Discover’s quality systems penalize clickbait (headlines that overpromise vs. content that underdelivers). Specific, informative headlines that convey genuine value outperform vague teasing headlines.
Authoritative topics within established interest areas: Content from sites with strong E-E-A-T signals in their topic area is surfaced more readily. A site known for technical SEO content is more likely to have its SEO articles surfaced to users with SEO interests.
Deep, engaging content: Discover rewards content that keeps users engaged after the click — long reads, visual content, video. Thin content that drives quick exits performs poorly.
Strong visual presentation: Beyond the required large image, visual-heavy content (infographics, photo essays, image-rich articles) performs better in the visual feed format.
Content That Underperforms in Discover
Pure evergreen reference content: How-to guides and reference articles without a timely angle tend not to be surfaced in Discover, which prefers recent content. The same article published when its topic is trending performs better than when published quietly.
Thin or text-only content: Articles with minimal visual assets below the image threshold, or very short articles with limited depth.
Keyword-optimized but boring titles: Titles optimized for search queries (“How to Fix Crawl Budget Issues: A Complete Guide”) can be less compelling as Discover cards than interest-driven titles (“Why Your Site’s Crawl Budget Is Being Wasted on Pages Nobody Reads”).
Measuring Discover Performance
Google Search Console has a dedicated Discover report showing:
- Total clicks from Discover
- Total impressions in Discover cards
- CTR per page
Key interpretation notes:
- GSC shows Discover data when there are enough data points (privacy threshold) — low-volume sites may see gaps
- Discover traffic is reported separately from web search traffic in GSC
- Article-level data in the Discover report shows which content resonates with Discover’s audience
Track CTR trends: Articles with high Discover CTR (the percentage of users who clicked a card after seeing it) tell you which headlines and topics connect with Discover’s audience for your site.
Compare content types: Over time, patterns emerge — what topics, formats, or article types generate Discover impressions. Use this to inform the editorial calendar.
The Discover vs. Search Trade-off
Discover optimization and search optimization point in different directions in some areas:
| Signal | Search | Discover |
|---|---|---|
| Keywords | Critical | Not relevant |
| Freshness | Minor (topic-dependent) | Major |
| Headlines | Query-aligned | Interest-driven |
| Content depth | Comprehensive | Engaging |
| Images | Supporting | Required/central |
Sites that publish both timely content (for Discover) and evergreen content (for search) capture both traffic sources. An editorial calendar that mixes reference content with timely takes on trending topics in the site’s area of expertise is a common approach to capturing both.