Official statement
Other statements from this video 9 ▾
- 2:06 Google adapte-t-il vraiment ses algorithmes en temps de crise ?
- 4:43 Le DMCA suffit-il vraiment à protéger votre contenu volé du duplicate content ?
- 8:30 Faut-il vraiment placer le balisage schema.org publisher sur toutes les pages de votre site ?
- 18:29 Le JavaScript peut-il transformer vos pages uniques en contenu dupliqué aux yeux de Google ?
- 20:44 Google lit-il vraiment le contenu des images pour les classer ?
- 36:11 Faut-il vraiment s'inquiéter des erreurs 404 qui s'accumulent dans la Search Console ?
- 39:23 Le contenu masqué en mobile-first est-il vraiment pris en compte par Google pour l'indexation ?
- 39:49 Les liens no-follow sont-ils vraiment ignorés par Google pour le crawl ?
- 41:52 Les données structurées profitent-elles au SEO même sans rich snippets visibles ?
Google requires images to be at least 1200 pixels wide to be eligible for Discover. This technical requirement comes with the use of the max-image-preview:large meta tag to enable the display of large thumbnails, even on non-AMP sites. Essentially, it is an absolute prerequisite — without these criteria, your content remains invisible in this recommendation feed, regardless of its editorial quality.
What you need to understand
Why does Google impose this minimum size requirement?
Discover is not a traditional search engine. It is a visual recommendation feed designed for mobile, where the image does 70% of the hook work. Google wants to avoid pixelated or too-small thumbnails that disrupt the user experience.
The 1200 pixels wide threshold ensures decent display quality on all screens, including high-resolution displays. Below this threshold, your content can technically be indexed, but it will never be featured in Discover — even if your article is excellent.
What is the max-image-preview meta tag and why should it be configured?
This robots tag controls the maximum size of image previews that Google can display across its surfaces (search, Discover, Google News). Three possible values: none, standard, large.
With max-image-preview:large, you allow Google to display large thumbnails — without this tag, even a 2000px image will be truncated or ignored. It's an explicit signal: 'yes, you can show my visuals in large format.'
The strategic interest? This tag works without AMP. Previously, large images in Discover were reserved for AMP pages. Now, any site can benefit from this — as long as the meta tag is properly configured in the <head>.
Does this requirement apply to all images on the site or just the main image?
Google looks at the cover image — the one defined in your Open Graph tags, Schema.org (ImageObject), or the first significant image on the page. There's no need to convert all your illustrations to 1200px.
That said, if your CMS automatically generates multiple sizes (thumbnail, medium, large), make sure that the version served to Google is indeed the full resolution. Some WordPress themes or caching systems serve compressed versions by default — and that’s enough to exclude you from Discover.
- Minimum 1200 pixels wide for the main image of each article eligible for Discover
- Max-image-preview:large robots meta tag in the
<head>of all relevant pages - No need for AMP to benefit from large thumbnails in Discover
- Check the version served to Googlebot — not the one displayed visually in the browser
- Only one image matters: the one defined as the cover image or the first significant image
SEO Expert opinion
Is this statement consistent with field observations?
Yes, and it’s one of the few Discover criteria that can be validated empirically. Sites that switch to 1200px+ see almost immediate visibility gains in Discover, provided the rest is in place (fresh content, engagement, eligible topic).
However, Mueller remains vague on a crucial point: width alone isn’t enough. It’s observed that Google also favors images with a 16:9 or 4:3 ratio, and a 1200×400 (very panoramic) image performs worse than a 1200×800. [To check]: Google does not communicate any official ratios, but field data points towards classic landscape, not banners.
What nuances should be considered regarding the max-image-preview meta tag?
The tag works, but be careful: it does not force Google to display your images in large format. It allows it. If your content doesn’t generate engagement, if the topic is beyond Discover's scope (very technical B2B, pure transactional content), the tag won’t change anything.
Another pitfall: some sites use overly restrictive robots.txt or inadvertently block Googlebot-Image. The result: the tag is present, the image is 1500px, but Google cannot crawl it. Always check in Search Console that your images are properly indexed.
In what cases does this rule not really apply?
If your SEO strategy does not target Discover — pure e-commerce site, B2B SaaS, local directory — this requirement is off-topic. Discover almost exclusively indexes fresh editorial content: news, lifestyle, sports, mainstream tech.
Even within eligible niches, presence in Discover remains unpredictable and volatile. You can check all the boxes (1200px, tag, quality content) and still not appear — Google guarantees nothing. Conversely, some sites benefit from massive boosts without understanding why. [To check]: the Discover algorithm incorporates engagement signals (CTR, time spent, shares) that we cannot directly control.
Practical impact and recommendations
What should be done concretely to optimize images for Discover?
Start with an audit of the main images of your recent articles. Extract all Open Graph and Schema.org images using Screaming Frog or a Python script, and filter those under 1200px wide. Prioritize evergreen content or articles that already generate SEO traffic.
Next, add the max-image-preview:large meta tag to your <head> template. On WordPress, plugins like Yoast or RankMath manage it automatically — just ensure the option is enabled. If you are custom, manually insert: <meta name="robots" content="max-image-preview:large">.
What mistakes should be avoided during compliance?
The first classic error: resizing the image displayed in CSS thinking it’s enough. Google looks at the source file, not the visual rendering. If your HTML loads an 800px image and the CSS scales it to 1200px, it does not count.
The second pitfall: malconfigured lazy loading or CDN systems. Some CDNs serve compressed or resized versions based on user-agent. If Googlebot receives a 600px version while visitors see a 1200px version, you’re out of the game. Test with the URL inspection tool in Search Console.
The third error: forgetting about duplicate images on mobile. Some themes serve different desktop/mobile images via <picture> or srcset. Ensure that the mobile version (the one Googlebot mobile sees first) is indeed at least 1200px.
How can I verify that my site is compliant and eligible?
Use the URL inspection tool in Search Console: paste an article URL, click on 'Test URL live', then inspect the screenshot. Check that the cover image is loaded and that it is at least 1200px wide.
Next, control the robots meta tag in the HTML source code returned to Google (not the one in your browser — use 'View crawled source'). If max-image-preview:large is absent or placed after a max-image-preview:standard, the latter takes precedence.
Finally, monitor the Discover performance reports in Search Console (under the 'Discover' section). If your articles never appear despite compliant images, the issue lies elsewhere: topic beyond scope, insufficient freshness, or too low engagement. These technical optimizations remain complex to orchestrate, especially if your CMS or technical stack adds multiple layers (CDN, cache, lazy loading, responsive images). Engaging a specialized SEO agency can be wise to finely audit your setup and identify invisible blocking points — an expert can cross-reference Search Console data, server logs, and Googlebot tests to precisely diagnose what’s causing the problem.
- Main images audit: identify those under 1200px wide
- Add the
max-image-preview:largemeta tag in the<head> - Check the source file served to Googlebot (not just the CSS rendering)
- Test via the URL inspection tool in Search Console
- Control CDN and lazy loading configuration (no reduced version served to the bot)
- Monitor Discover reports in Search Console
❓ Frequently Asked Questions
Puis-je utiliser des images de 1200px en hauteur plutôt qu'en largeur ?
Le méta-tag max-image-preview:large impacte-t-il aussi la recherche classique ou seulement Discover ?
Faut-il modifier les anciennes images ou seulement les nouvelles publications ?
Est-ce que les formats WebP ou AVIF posent problème pour le critère de 1200px ?
Si j'ai plusieurs images dans un article, laquelle Google utilise-t-il pour Discover ?
🎥 From the same video 9
Other SEO insights extracted from this same Google Search Central video · duration 54 min · published on 31/03/2020
🎥 Watch the full video on YouTube →
💬 Comments (0)
Be the first to comment.