--- title: "How Do I Get Into Google Discover?" author: "Roey Granot" category: "AI-Transformed Workflows" date: 2026-09-26T11:00:22.983Z canonical: "https://contentagents.dev/blog/how-do-i-get-into-google-discover-ad51" --- # How Do I Get Into Google Discover? ![Hands holding a smartphone displaying a scrolling card feed, mid-swipe, with a half-drunk coffee on a wooden table in morning](https://hsppuvezyxmkpzkgfkho.supabase.co/storage/v1/object/public/media/enrichment/024a6468-4c4c-4195-b8c2-21b4170617d4/79e6ac9a-b83d-489a-8a8c-5241ebc0796c/51c74dfc-7676-4d74-ae0c-31a8cb3b6869.jpg) Google Discover can send thousands of readers to an article they never searched for - and most content teams have no idea why it works when it does. This guide gives you the six-step process that determines whether Google surfaces your content in Discover feeds, based on how the algorithm actually behaves. This is written for founders and content operators who publish regularly but haven't cracked Discover yet. It assumes you're already indexed by Google and publishing at least a few articles a month. If you're starting from zero, build your content base first, then return here. A note on what we know directly: we work with content programs every day and see firsthand which publishing behaviors correlate with Discover traction. Where we're synthesizing research from others - particularly on algorithm mechanics - we say so explicitly. ## Step 1: Build topical authority in a specific niche before expecting Discover traffic ![](https://cdn.pixabay.com/photo/2018/01/24/21/42/step-3104846_1280.jpg?w=960&q=75) Photo by [mikecook1](https://pixabay.com/photos/step-industry-steel-3104846/) on [Pixabay](https://pixabay.com) Discover surfaces content from sources it recognizes as authoritative in narrow topics. A finance blog that publishes only about retirement planning for self-employed workers will outperform a general business blog, even if the general blog has more total traffic. Google's systems learn your coverage pattern over time - consistent deep coverage of one niche trains the algorithm to trust your voice on that subject. This is different from traditional SEO, where you target keywords and pick up rankings across many topics. In SEO, breadth can be a strategy. In Discover, breadth is a liability. The algorithm is trying to predict what a specific reader will want to read next - and it can only make that prediction confidently if it understands what your site covers. A measurable way to think about this: aim for 50 or more articles on the same topic, published over at least six months, with internal linking that shows conceptual depth - not just a list of posts with the same tag. Avoid covering the same topic once every few months with unrelated content in between. Consistency is what creates the signal. ### What this means in practice - Audit your content library. If your top 20 articles span eight or more different topics, you're not ready for Discover yet - and that's useful information, not a failure. - Pick one niche and commit to it for at least three months before expecting Discover impressions to grow. - Internal linking matters here. Link articles on related subtopics to each other so Google can map the conceptual territory you're covering. - Breadth kills the signal. One article on email marketing, one on hiring, one on sales ops - Google can't build a coherent picture of what you're about. ## Step 2: Write headlines that trigger curiosity without manufacturing desperation Discover users are scrolling a feed, not typing a query. Your headline has roughly 1.5 seconds to answer "why should I care?" in under ten words. Curiosity-driven headlines that promise specific value outperform clickbait, which Discover's algorithm actively penalizes through low engagement signals. The practical difference looks like this: "Why Google Discover Traffic Spikes in the First 48 Hours" is a Discover headline - it creates a knowledge gap and implies a specific, useful answer. "You Won't Believe What Google Discover Does to Your Traffic!!" is clickbait - it creates expectation without substance, and users who tap and bounce teach the algorithm your content isn't worth surfacing. Discover headlines optimize for human curiosity. SEO headlines optimize for keywords. The same article can have both: your URL slug and meta title target the keyword phrase; your on-page headline and social copy target the curiosity angle. These don't have to conflict. ### What this means in practice - Check your headline against this: does it contain a number, a time frame, or a specific outcome? If not, it's probably too vague for Discover. - Avoid ALL CAPS, exclamation marks, and question marks that sound anxious. These are markers the algorithm associates with low-quality content. - Write two headline versions per article: one keyword-optimized for SEO, one curiosity-optimized for Discover. Use the Discover version in social promotion. - Read the headline out loud. If you'd feel embarrassed reading it to a colleague, rewrite it. ## Step 3: Use custom featured images at 1,200px wide minimum - stock photos cost you clicks Discover is a visual feed. Before a user reads your headline, they see your image. A blurry or generic stock photo tells the algorithm - through low tap rates - that your content isn't worth showing again. According to Search Engine Land's Discover guide, image quality is one of the most consistently cited factors in whether content surfaces and performs in the feed. The technical floor: 1,200px wide minimum, 16:9 aspect ratio preferred, file size under 200KB. But the specs aren't the real constraint - the content of the image is. Discover's algorithm learns from user behavior. When users tap on articles with specific, high-signal images - a custom chart, a branded graphic, an original photo - that teaches the system those images predict engagement. When they don't tap on generic stock photos, that teaches the opposite. Two articles on the same topic with the same headline quality will perform differently in Discover if one uses a stock photo of a person at a laptop and the other uses a custom chart showing the data the article discusses. The chart signals: this is original, specific, worth reading. ### What this means in practice - Audit your last ten published articles. How many featured images did you create or commission? If most are stock photos, that's your fastest fix. - Custom doesn't mean expensive. A branded Canva template with article-specific text and a chart can outperform a polished stock photo. - Mark your images as large enough for Discover in your CMS. In WordPress with Yoast or RankMath, enable the "large image" setting in your social/Discover options. - File size matters. A 1,400px image at 800KB will load slowly and depress engagement. Compress to under 200KB using a tool like Squoosh before uploading. ## Step 4: Include recent data published within the last 90 days - recency is a hard filter Discover prioritizes freshness. An article published six months ago with outdated statistics is unlikely to surface, even if the writing quality is high. The algorithm is designed to show users things happening now - which means your content needs to signal currency, not just quality. A concrete example: if you're writing "Marketing Trends in 2025" and you publish in January with data from December 2024, you catch the recency wave. If you publish the same article in April with the same December data, you've missed it. The data is now five months old, and the topic has passed its peak interest window. The working rule: include at least one statistic, study, or event published within the last 90 days and link to the original source. This signals to Google that you're tracking your topic actively, not recycling old content. It also gives users a reason to trust the article is current. Where we've seen this matter most - and we see it across content programs we work with - is in fast-moving topics like AI, marketing, and finance, where stale data is obvious to readers. If you're tracking content performance across multiple articles, [engineering a content intelligence system with GA4 and GSC](/blog/beyond-the-pageview-engineering-content-intelligence-with-ga4-gsc-8abd) can help you spot which pieces are losing their recency edge before it hurts your Discover impressions. ### What this means in practice - Before publishing, verify that your most important statistic has a publication date within the last 90 days. If not, find a newer source or update the claim. - Link to the original source, not a secondary article that cites it. Google can follow the chain, and so can readers - going to the secondary source looks lazy. - Set a calendar reminder to refresh high-performing Discover articles every 90 days. A small data update with a republish date can revive a piece that's sliding out of the feed. - Don't fake recency. Adding a current year to an old headline without updating the content is a pattern the algorithm - and readers - recognize. Bounce rates spike, and the signal turns negative. Publishing consistently at this cadence is harder without a structured workflow. Teams that [balance AI-assisted content production with human editorial review](/blog/ai-content-workflows-where-human-review-still-matters-jq5d) tend to maintain the output frequency Discover rewards without sacrificing the accuracy that recency-dependent articles require. ## Step 5: Earn engagement signals in the first 48 hours after publishing Discover operates on a feedback loop. When you publish, Google shows your article to a small test audience. If that audience engages - taps, reads, shares - Google expands distribution. If they don't, the article gets buried. This is why the first 48 hours after publishing are disproportionately important, and why driving your own initial traffic matters. The signals that matter: tap-through rate from the Discover feed, time on page, scroll depth, and low bounce rate back to the feed. These tell the algorithm whether the article delivered on its promise. A high tap-through rate combined with strong time on page is the combination that unlocks broader distribution. Your owned channels are your distribution lever here. Email lists, social accounts, and Slack communities can seed that initial engagement wave. If you have 2,000 email subscribers and 30% open the article in the first 24 hours, you've created the engagement signal Discover needs to test broader distribution. This is why content programs with distribution assets outperform those without, even at the same publishing frequency. ### What this means in practice - Schedule your email newsletter to go out within 24 hours of publishing. Don't wait for your weekly digest - Discover's test window doesn't align with weekly cadences. - Share in relevant communities the day of publishing. Not spam - a single, relevant post in a Slack channel or LinkedIn group where the article genuinely fits. - Monitor Search Console's Discover report daily for the first week after a new article goes live. If impressions don't appear within 48–72 hours, the article didn't pass the initial test. - Optimize for scroll depth, not just clicks. An article that gets tapped but abandoned at paragraph two is worse than no taps. This means your lede has to earn the read - don't bury the value. ## Step 6: Monitor Discover-specific metrics in Search Console and iterate on what works ![](https://cdn.pixabay.com/photo/2019/07/14/08/21/stands-4336430_1280.jpg?w=960&q=75) Photo by [dimitrisvetsikas1969](https://pixabay.com/photos/stands-steps-symmetry-theatre-4336430/) on [Pixabay](https://pixabay.com) Most teams monitor their Search Console Performance report for search traffic and ignore the Discover tab entirely. The Discover report shows impressions, clicks, and click-through rate from the feed - and the patterns in that data tell you which of your articles the algorithm chose to test, and which ones it distributed broadly. The metric to watch first isn't impressions - it's click-through rate. A high impression count with a low CTR means Google is showing your article but users aren't tapping. That's a headline or image problem. A low impression count means the article never passed the initial test - that's a topical authority or recency problem. Separating these two failure modes tells you where to focus. One pattern we see consistently: articles that perform well in Discover share a combination of features - a specific, curiosity-driven headline, a custom image, recent data, and strong initial engagement. When an article fails, it usually fails on one of these dimensions specifically. The Discover report, read alongside your on-page analytics, shows you which one. The [content ROI metrics most B2B teams get wrong](/blog/the-content-roi-metric-most-b2b-teams-get-wrong-vi4d) often include Discover performance - teams that only track last-click attribution miss the full picture of how feed-driven traffic compounds over time. ### What this means in practice - Open your Search Console Discover report weekly. Sort by impressions descending and look for articles with high impressions but low CTR - those are your headline and image optimization targets. - Build a simple log: article title, publish date, Discover impressions at 7 days, CTR. After ten articles, patterns will emerge about which topics and headline formats the algorithm favors for your site. - Don't delete or significantly rewrite articles that performed well in Discover. Changing the URL or headline after an article has built Discover history resets the signal. - Use the Discover data to inform your editorial calendar. If articles about one subtopic consistently generate Discover impressions, publish more in that subtopic - you're confirming the algorithm's topical model of your site. Getting into Google Discover can feel opaque, but the traffic potential when it works is hard to ignore. Jesse Cunningham has built a reputation for testing Discover aggressively, and his findings on what actually triggers inclusion are worth examining alongside the strategic advice in this article. ## FAQ ### How long does it take to start getting traffic from Google Discover? Expect a realistic timeline of three to six months if you run all six steps consistently. The first three months are spent building topical authority - publishing regularly in one niche with strong internal linking. From month four onward, running the full process on every article (custom image, recent data, cross-channel promotion within 24 hours, weekly GSC review) gives the algorithm enough signal to start surfacing your content. ### How many articles do I need before Google Discover will pick up my content? Aim for 50 or more articles on the same topic, published over at least six months, with internal linking that shows conceptual depth. If your top 20 articles span eight or more different topics, you are not yet sending a clear enough topical authority signal for Discover to recognize your site as a trusted source in any specific niche. ### What image size and format does Google Discover require? Your featured image should be at least 1,200px wide, ideally in a 16:9 aspect ratio, and compressed to under 200KB. Beyond the specs, the content of the image matters - custom charts, branded graphics, or original photos outperform generic stock photos because users tap on them more often, which teaches the algorithm your content is worth surfacing. ### How do I check if my articles are appearing in Google Discover? Log into Google Search Console, click 'Performance,' and filter by 'Discover.' You will see impressions, clicks, and click-through rate for each article. A healthy Discover CTR is roughly 3 to 8 percent. If impressions are climbing but CTR is flat, your headline or image is likely the problem. If impressions are low after two months, revisit your topical authority - that is the most common root cause of zero Discover traffic. ### Does promoting a new article really affect whether Google Discover picks it up? Yes - promotion within the first 24 hours is a key part of how Discover decides whether to surface your content more broadly. The algorithm learns from user behavior like clicks, time on page, and shares. An article that gets 500 clicks in the first 48 hours through email, LinkedIn, and other channels gives Google behavioral data to act on. An article that gets 30 clicks because it was not promoted gives the algorithm no reason to surface it further, regardless of writing quality. --- Source: https://contentagents.dev/blog/how-do-i-get-into-google-discover-ad51