Does Discover really have no exploitable eligibility criteria?

Does Discover truly operate without any query logic?

Yes, and this is precisely what distinguishes it from the traditional SERP. Discover proactively pushes content without the user formulating an explicit intention. The algorithm predicts what might interest the user based on their history, behavioral signals, and declared or inferred interests.

Whereas a traditional Google query relies on a matching intent-content, Discover functions on a probabilistic prediction of interest. As a result: no keyword to target, no search volume to track, no SERP to analyze. The playing field is radically different.

Why does Google refuse to provide clear criteria?

Because revealing eligibility criteria would open the door to mechanical optimization that would degrade the quality of the stream. If Google said, "you need X backlinks + Y click-through rate + Z freshness," you would see content farms industrialize the process. Discover thrives on its ability to surprise and engage — not on becoming a predictable channel.

Another reason: the criteria likely vary by user profile. The same article may appear for a tech enthusiast and never for a general reader. It's difficult to formulate a universal rule when the engine personalizes to such an extent.

What really differentiates Discover from traditional search?

In traditional search, you have an implicit guarantee of indexing: if your content is crawled, technically sound, and relevant to a query, it has a chance of appearing somewhere in the results. Even in position 87, it exists.

With Discover, there's no guarantee of appearance. Your content can be technically perfect, editorially solid, from an authoritative site — and never hit the stream. Google can decide to reduce an entire site's visibility without explanation or boost it massively overnight. Volatility is structural.

• Discover does not respond to a query: intent is predicted, not formulated.
• No public criteria: Google can promote or reduce a site without transparency.
• No guarantee of appearance: unlike traditional search where indexing = potential eligibility.
• High volatility: traffic can explode or collapse without any editorial changes on your part.
• Extreme personalization: the same content can appear for some users and never for others.

Is this statement consistent with field observations?

Yes, unfortunately. SEOs who have been tracking Discover for years observe a total absence of stable correlation between optimizations and results. A site can go from 0 to 500K Discover sessions in a week without having changed anything, and then drop to zero two months later. No reproducible pattern.

Some factors still seem to carry weight — high-quality images, emotional titles, content freshness, user engagement — but their impact remains probabilistic, not deterministic. You can check all the boxes and never appear. Conversely, an average article can explode for no apparent reason. [To be verified]: Google claims there are no criteria, but some qualitative signals still seem to play a role — the question is which ones and to what degree.

In which cases does this rule not apply?

Let's be honest: some publishers seem to have a structural privileged access. Major media outlets, established news sites, domains with a strong history in Discover enjoy near-guaranteed visibility. Not because they optimize better, but because their editorial authority gives them a de facto algorithmic boost.

If you're Le Monde, CNN, or The Guardian, you don't have to wonder if you'll appear in Discover — you're there by default. For a small site, even an excellent one, it's a lottery. Mueller's statement is true for the majority of sites, but asymmetric according to editorial authority.

What nuances should be added to this statement?

Mueller says "no guaranteed method," which is true. But he doesn't say "no exploitable signals." In practice, some levers still seem to correlate with an increased probability of appearance: large images (min 1200px), solid AMP or Core Web Vitals, high user engagement, evergreen or ultra-fresh content depending on the sector.

The problem is that these signals guarantee nothing — they just increase your chances. It's like playing poker: you can maximize your edge, but you never control the final outcome. Optimization for Discover remains a gamble, not a science.

What should you do concretely if you want to maximize your chances?

First rule: never rely on Discover as your main channel. It's a bonus, not a strategy. If your business depends on this traffic, you are in danger. That said, some levers statistically increase your chances of appearing.

Favor very high-quality images (at least 1200px wide), with a 16:9 or 4:3 ratio. Generic or too-small visuals seem filtered. Work on your titles to generate an emotional or cognitive reaction — curiosity, surprise, immediate usefulness. Discover favors content that quickly engages.

Ensure your pages load quickly and seamlessly. Core Web Vitals likely matter less than in traditional SERP, but a slow site reduces engagement — and Discover seems to incorporate post-click behavioral signals. If users bounce immediately, your content risks being downranked.

What mistakes should be absolutely avoided?

Don't fall into the trap of mechanical optimization. Over-optimizing for Discover can penalize you elsewhere — particularly in classic SEO if you distort your titles or content to chase emotion at the expense of relevance. The game rarely pays off.

Also avoid obsessively tracking Discover. Traffic can explode one day, disappear the next, without you having done anything. If you react to every fluctuation, you will lose yourself. It's better to focus on predictable and measurable channels.

How can I check if my site is at least eligible?

Check in Search Console > Performance > Discover. If you have impressions, even minimal, it means Google at least considers some of your pages as eligible. If you have none at all, either your content doesn't match any user profile, or your site lacks sufficient authority to enter the stream.

Test by creating visually rich, emotionally engaging, and ultra-fresh content on a trending topic. If even that generates nothing, it's probably an authority problem — and there's no quick solution. Optimization for Discover requires a deep understanding of user signals and continuous editorial adjustments.

• Never be structurally dependent on Discover traffic
• Favor high-quality images (min 1200px)
• Craft titles to generate emotion or curiosity
• Optimize speed and Core Web Vitals
• Track Discover in Search Console without overreacting to fluctuations
• Test visual + trending content to check eligibility