How does Google determine which image to display in local search results?

Why does this statement specifically relate to local results?

Local results sliders appear in SERPs when the search intent includes a geographical dimension — restaurant near me, plumber Paris 15, hairdresser Lyon. These carousels showcase a highlighted image for each establishment, creating a visually impactful element on the click-through rate.

Mueller clarifies that Google does not randomly pick: it identifies the main image of the page by cross-referencing structured data and visual analysis. This approach differs from conventional images in traditional organic results, where Google can draw from all visual content on the site. Here, it's a strategic selection to represent the local entity.

What does "structured data and accessible visual content" mean in this context?

Google refers to Schema.org markup, particularly the types LocalBusiness, Restaurant, Organization and their variants. The 'image' property of these schemas explicitly indicates which photo represents the establishment. It is the most direct signal that can be sent to Google.

"Accessible visual content" refers to images present in the HTML of the page, their position in the DOM, their actual size, and their semantic context. An image at the top of the page, in the hero area with relevant alt text, will carry more weight than a thumbnail lost in the footer. Google also analyzes the perceived visual weight — a large storefront photo takes precedence over a small logo.

Which types of pages are impacted by this mechanism?

Primarily establishment pages: local listings, landing pages for physical agencies, "Find Us" pages, one-page business sites. Any URL that represents an entity with a physical address and aims to rank for local queries falls within this scope.

Multi-establishment sites (chains, franchises, agency networks) must apply this logic to each location page. A common mistake is to markup only the corporate homepage, leaving local pages orphaned of structured markup. The result: Google picks a generic or inappropriate image.

• Prioritize LocalBusiness markup or its sub-types (Restaurant, Store, HealthAndBeautyBusiness…) with an explicit 'image' property
• Place the strategic image at the top of the page, visible without scrolling, in a format suitable for carousels (minimum 1200x630px, ratio 16:9 or 1:1)
• Optimize technical weight: descriptive alt text, meaningful filename, modern WebP or JPEG format, lazy-loading disabled for the main image
• Test with the structured data testing tool and Search Console to ensure Google correctly recognizes the marked image
• Avoid images with embedded text, standalone logos, or generic visuals — prefer an authentic photo of the establishment, flagship product, or team

Does this statement introduce new elements or confirm known practices?

Let's be honest: this statement confirms what has been observed in the field for several years, without truly unveiling the specific prioritization criteria. It was already known that Schema.org influences the selection of images in rich results. What is lacking here is transparency regarding conflict cases.

What happens when multiple images are marked in the same markup, or when the Schema image differs from the Open Graph image or that uploaded in the Google Business Profile? Mueller does not address this. Field tests show that Google often prioritizes the Schema image if it meets technical criteria, but not always. [To be verified] with A/B tests on your own local pages.

What limitations does this automatic approach pose for SEO professionals?

The automatic selection means a total loss of control. Even with perfect markup, Google may decide that another image on the page is more "representative" according to its computer vision algorithms. I have seen cases where an event image was preferred over the permanent storefront photo, simply because it was larger and better contrasted.

For brands with a strict visual identity, this is a risk. The image displayed in local sliders may not align with the desired visual identity. The only workaround: run multiple tests, regularly monitor local SERPs (in private browsing, from different geolocations), and adjust markup according to observed results. Specifically? A monthly visual audit of your clients' top local queries.

In which cases does this rule not apply or yield unexpected results?

I have observed erratic behaviors on multilingual pages or sites with several regional versions. Google may display the image from the en-US version in fr-FR results if the hreflang is misconfigured or if the marked image is not accessible from the local version. Crawling and rendering must be flawless.

Another problematic case: sites with aggressively lazy-loaded images. If the main image is not in the initial HTML but injected by JavaScript after user interaction, Google may not see it during rendering and choose a poor fallback. Disable lazy-loading for the hero image of your local pages — the gain of a few milliseconds is not worth the risk of missing the display in sliders.

What concrete steps should be taken to optimize the image displayed in local results?

Start with an audit of your establishment pages. List all URLs representing physical locations and check the existing Schema markup. Use Google's structured data testing tool to identify pages lacking LocalBusiness markup or missing 'image' property.

Next, select for each page the most strategic image: storefront, flagship product, team, welcoming interior. Avoid generic visuals, standalone logos, or images with embedded text (not accessible, poor for mobile UX). Optimize this image: modern format (WebP), appropriate dimensions (minimum 1200x630px), weight < 200Ko, descriptive alt text including the establishment name and city.

What mistakes should be absolutely avoided in the implementation?

Do not markup multiple competing images in the same LocalBusiness Schema without a clear hierarchy. If you need to add a gallery, use a separate secondary images array, but explicitly designate the main image with the singular 'image' property, not an array.

Avoid copy-pasting the same markup on all pages of a multi-establishment network. Each location must have its own markup with its specific image, unique address, and dedicated phone number. Generic markup dilutes local relevance, and Google may ignore signals or display an inappropriate image.

How can I verify that my optimization is working and measure the impact?

After deploying markup, wait 2-4 weeks for Google to recrawl and reprocess your pages. Then perform local searches from different geographical positions (tools like BrightLocal, Local Falcon, or private browsing with VPN). Capture screenshots of sliders to check which image is displayed.

Monitor the local organic click-through rate in Search Console by filtering for geolocated queries. A more engaging image should improve CTR for local results. Compare before/after over a period of at least 60 days to smooth seasonal variations. If you manage multiple establishments, test first on a pilot sample before deploying at scale.

• Implement Schema.org LocalBusiness (or appropriate subtype) with explicit 'image' property on each establishment page
• Place the strategic image in the hero position, visible without scrolling, minimum format 1200x630px, ratio 16:9 or 1:1
• Technically optimize: modern WebP/JPEG, < 200Ko, descriptive alt text, no lazy-loading on the main image
• Validate the markup with Google’s testing tool and check for errors in Search Console
• Monthly monitor local SERPs (screenshots, CTR in Search Console) to detect any display changes
• Audit CDN images and check CORS headers to avoid crawl blocks