A client asked me last week why their Google Analytics showed a traffic spike from something called "Discover" that had nothing to do with any keyword they were targeting. That's a fair question. Most business owners have never heard of Discover as its own thing, let alone read what Google actually publishes about getting into it. Since it directly follows the February 2026 Discover update, I went and read that documentation this week. Here's what it actually says, not the guessed-at version most SEO blogs repeat.

How Discover Actually Works Inside Google Search

Discover is part of Google Search, but it works nothing like typing a query. Google's own description: it's "a part of Google Search that shows people content related to their interests, based on their Web and App Activity." Someone opens the Google app or a new Chrome tab on their phone and sees a feed, no search box involved. Ranking for "plumber Tulsa" and showing up in someone's Discover feed are two separate systems, built on different signals, and doing well at one doesn't guarantee anything about the other.

The Bar to Clear, According to Google

According to Google's Get on Discover documentation, content is automatically eligible if it's indexed and meets Discover's content policies, with no special tags or structured data required. Eligibility isn't a guarantee of appearing, though, and that gap is where the actual guidance lives: avoid clickbait and misleading titles, avoid sensationalism, and provide content that's timely or offers a real, unique insight rather than a rehash. The February 2026 update added one more layer on top of that: Google now judges expertise topic by topic, not site-wide, so a site with a genuinely deep section on one subject gets credit there even if the rest of the site covers something unrelated.

The Image Spec Almost Nobody Follows

This is the part that's actually actionable and mostly ignored. Google's specifications: images at least 1200 pixels wide, with more than 300,000 total pixels, in a 16:9 aspect ratio, specified through the og:image tag or schema markup, with max-image-preview:large set to allow the large format. Google also recommends avoiding generic images like a site logo and avoiding text-heavy graphics in that slot, since Discover tends to favor a real, representative photo over a designed card. This site already sets max-image-preview:large sitewide and ships a properly sized og:image on every post, precisely because of this documentation, not as a guess.

It Also Remembers What Someone Already Reads

One line in Google's announcement is easy to skip past: Discover will "continue to show content that's personalized based on people's creator and source preferences." That's a separate mechanism from topic expertise. Someone who regularly reads a specific local news site, or clicks through to the same handful of creators, is more likely to keep seeing content from those same sources, independent of how deep any single article is. In practice, that means Discover rewards consistency over a single strong post. A business that publishes once a year and happens to nail the topic depth is still starting from zero with any given reader. A business that shows up regularly has a better shot at becoming one of those recognized sources.

Who This Actually Helps

Discover is built to present interest-based content, mainly articles, and it filters out things like job applications, petitions, forms, and code repositories. A site with a homepage, three service pages, and a contact form has almost nothing shaped like what Discover is looking for, since there's no article there to surface. That's exactly why keeping a genuine, regularly updated knowledge base matters more than any single update, Discover eligibility depends on having actual article-shaped content to begin with.

Where Discover Traffic Actually Fits

Google is direct about this too: "traffic from Discover is less predictable or dependable when compared to keyword-driven search visits," and it recommends treating it as supplemental to keyword-driven traffic rather than a replacement for it. A Discover spike works best as a nice surprise sitting on top of the work that already ranks a business for what people actually type into Google, and chasing it instead of that underlying work tends to leave both undone.