How can you optimize for Google Discover without targeting keywords?

Does Discover really work differently from traditional search?

Yes, and this is precisely what confuses many SEO practitioners. Google Discover does not respond to an explicit search intention formulated by the user. It proactively presents content based on a dynamically constructed interests profile drawn from browsing history, past searches, and interactions with other content.

Mueller's statement highlights a crucial point: traditional queries take a back seat in this context. Discover prioritizes overall thematic relevance, editorial authority on a given topic, and a content's ability to generate engagement. This is a paradigm shift for those used to calibrating everything around keyword research.

What replaces keyword optimization then?

Mueller talks about monitoring topics that resonate with users. Practically, this means analyzing engagement metrics — time spent, bounce rate, scroll depth, social shares — to identify themes that truly perform. The Discover algorithm continuously learns from these behavioral signals.

The approach becomes iterative: publishing diverse content on adjacent topics, observing what generates the most engagement, and then adjusting the editorial line accordingly. It’s a test & learn logic rather than query targeting. Publishers need to act more like media outlets than traditional SEO sites.

Can relevance and engagement be measured concretely?

Google remains deliberately vague about the exact metrics used to evaluate relevance in Discover. It can be assumed that E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) plays a significant role, as well as content freshness and the historical engagement of previous publications.

On the engagement side, classic signals (CTR, session duration, interactions) are likely taken into account, but Google does not communicate thresholds. The practitioner thus must rely on Google Search Console — Discover section — to analyze impressions, clicks, and CTR by URL, and cross-reference this data with their analytics to understand performance patterns.

• Discover does not respond to an explicit request but to a dynamically constructed user interests profile
• Traditional keyword optimization becomes secondary in favor of overall thematic relevance
• Continuous performance monitoring by topic is essential for adjusting the editorial strategy
• Engagement metrics (time spent, scroll depth, interactions) become central KPIs
• E-E-A-T and content freshness likely weigh heavily in the Discover algorithm

Is this statement consistent with what we observe on the ground?

Overall, yes — but with important nuances. Sites that perform well in Discover indeed share strong editorial characteristics: consistent editorial line, regular publication, quality visual content (high-resolution images, videos). It is also observed that news sites, lifestyle blogs, and niche media dominate, confirming that Discover favors players who publish fresh and engaging content.

However, to say that keyword optimization has no importance is reductive. Even in Discover, image alt text, H1 titles, and the semantic structure of content likely influence Google’s ability to understand the subject matter. One cannot completely ignore on-page SEO — simply put, it is no longer sufficient. [To be verified]: Google has never detailed the respective weight of semantic versus behavioral signals in Discover.

What are the practical limits of this editorial approach?

The main issue is that monitoring topics that resonate requires a sufficient initial traffic volume to draw statistically valid conclusions. A small site just starting out will not have enough Discover data to iterate effectively — a classic vicious circle.

Moreover, Discover's volatility presents a real challenge. Content can surge in impressions for 48 hours and then drop to zero for no apparent reason. This unpredictability complicates any medium-term editorial planning. Publishers must accept a degree of uncertainty incompatible with traditional SEO forecasts. Lastly, the dependence on engagement signals can lead to clickbait content at the expense of depth — a slippery slope.

In what cases does this strategy not apply?

Discover is simply not relevant for all types of sites. Transactional sites (pure e-commerce, SaaS, B2B lead generation) generally have no leverage on Discover, which massively favors informational content and news. If your business model relies on converting a specific search intention, investing in Discover is likely a waste of time.

Similarly, highly regulated YMYL (Your Money Your Life) sectors — health, finance — struggle more to perform in Discover, where Google seems to apply even stricter quality filters. Finally, very technical evergreen content often performs better in traditional organic search than in Discover, which prioritizes freshness and timeliness. Let’s be honest: Discover is a complementary channel, never a stand-alone SEO strategy.

How can you concretely identify topics that resonate with your audience?

The first step is to activate and regularly monitor the Discover section in Google Search Console. It provides impressions, clicks, and CTR by URL — essential data for spotting performing content. Cross-reference this data with your analytics to analyze post-click engagement metrics: session duration, pages viewed per visit, adjusted bounce rate.

Next, segment your content by theme or editorial category, and compare performances. Some topics may generate many impressions but few clicks — a sign of a title or image issue. Others may have an excellent CTR but low engagement — the content doesn’t fulfill the promise of the title. This analysis allows you to refine your editorial line in a data-driven rather than intuitive manner.

What mistakes should be avoided in a Discover-oriented strategy?

A classic mistake is producing content solely for Discover while neglecting traditional SEO. This puts all your eggs in a highly volatile basket. Discover should be a complementary channel, never the sole focus. Another trap is sacrificing depth and quality for catchy titles and superficial content. This may work in the short term, but Google regularly adjusts its quality filters — and your site could get booted.

Be careful not to over-optimize visually at the expense of text. Yes, high-resolution images are crucial for Discover, but content that lacks substance will not retain users. Finally, many sites publish irregularly — this is a deal-breaker for Discover, which favors active and consistent publishers. A predictable publication rhythm (minimum daily or weekly) is almost mandatory.

What needs to be set up technically and editorially?

On the technical side, ensure your site meets Google's quality content requirements: AMP optional but recommended, images must be at least 1200px wide, meta tags should be properly filled (especially Open Graph for previews). Also check that your RSS feed is clean and that structured data Article is correctly implemented on each publication.

On the editorial side, build a matrix of diverse content on topics related to your area of expertise, publish regularly, and iterate based on Search Console data. Develop a strong editorial identity: your angle, tone, and visuals should be recognizable. Finally, encourage user engagement — comments, shares, reading time — by structuring your articles in a scannable manner and incorporating multimedia.

• Activate and weekly monitor the Discover section in Google Search Console
• Cross-reference GSC data with analytics engagement metrics (duration, scroll depth, bounce)
• Segment content by theme to identify high-potential topics
• Meet technical requirements: images 1200px+, structured data Article, clean RSS feed
• Publish regularly (minimum weekly) with a coherent editorial line
• Avoid pure clickbait — prioritize sustainable engagement over fleeting virality