Image SEO is the process of preparing photos and graphics so search engines can read them, load them quickly, and rank them in visual search tools like Google Images and Google Lens. Two things changed in 2026 that most existing image SEO guidance has not caught up to: FAQ rich results were removed from Google Search, and Google raised the minimum resolution requirement for product images. Both are covered below with their sources and dates.
Why Image Search Is Worth Optimizing For
Search is not limited to typed queries. Google Lens lets a searcher photograph a product, a part, or a storefront and returns visual matches, meaning a business with unoptimized images is simply absent from that search path regardless of how strong its text content is.
For industries with strong visual inventory, manufacturing, retail, and food among them, Google also displays an image row above the standard text results for many queries. A business ranking there is visible before the first text result loads.
The Image Optimization Workflow
1. Choose the right file format
JPEG suits standard photos with many colors. PNG suits transparent backgrounds and simple logos. WebP compresses further than either without a visible quality loss, and Google’s own documentation recommends it for general web use, which makes it the default choice for most site images.
2. Resize before you upload
An image should be uploaded at the size it will actually display, not resized by the browser after the fact. A blog display area 800 pixels wide should receive an 800-pixel-wide image, not a 4,000-pixel original. Largest Contentful Paint, one of Google’s Core Web Vitals metrics, is directly affected by unnecessarily large image files.
3. Name the file with real context
Google’s image-understanding systems read filenames as a relevance signal. DSC_00982.jpg tells a crawler nothing. industrial-cnc-lathe-florida.jpg tells it what the image contains and where it’s relevant. Use hyphens between words, not underscores.
4. Write alt text that actually describes the image
Alt text serves two purposes at once: it’s read aloud by screen readers for accessibility, and it’s a ranking input for Google Images. “Machine” describes nothing. “High-speed 5-axis CNC machine processing aluminum parts” describes the image the way a person would over the phone. Keyword-stuffed alt text serves neither purpose.
5. Place images next to the text they illustrate
An image of a specific part or product should sit next to the paragraph discussing that part, not in an unrelated gallery. Google weighs an image’s surrounding text as context for what the image shows.
6. Understand what structured data actually does for images now
FAQ rich results (the expandable question dropdown in search results) have been narrowing for years. Google first restricted them to well-known government and health sites in August 2023. In June 2025, Google separately retired seven other structured data types for the same reason: low usage, little user value. On May 7, 2026, Google added a deprecation notice to its official FAQPage documentation confirming FAQ rich results no longer appear in Google Search at all, ending the feature even for the government and health sites that had kept it. The FAQPage schema type itself is still valid and won’t cause errors, it simply no longer produces that visual feature for anyone.
What structured data still does, and arguably does more than before: Google’s Gemini-powered AI Mode reads schema markup as a trust and entity-verification signal during answer synthesis, not primarily as a display trigger. Accurately describing a product, business, or image through structured data increases the odds of AI Mode citing that content, even with no traditional rich result attached. For product and catalog images specifically, Product structured data remains one of the highest-value implementations available.
7. Build an image sitemap for JavaScript-loaded galleries
Images loaded dynamically through JavaScript, common in product galleries, are not guaranteed to be discovered by a standard crawl. A dedicated image sitemap gives Google an explicit list of every image worth indexing.
Image SEO for Product and Industrial Catalogs
Catalog and inventory photography carries a different set of problems than a typical blog image, and most general image SEO guidance does not address them, particularly one specific 2026 change.
Google raised the minimum resolution requirement for Merchant Center product images to 500×500 pixels. Warnings for images below this threshold began appearing in Merchant Center accounts on April 14, 2026, with full enforcement starting January 31, 2027 (source: Google Merchant Center Help documentation). Google has said it may temporarily optimize some smaller images using AI upscaling to prevent disapproval in the interim, but that shouldn’t be treated as a substitute for fixing the source catalog. Any business running Shopping campaigns or free product listings on a manufacturing or industrial catalog should audit existing product images against this threshold before the enforcement date, not after.
Beyond resolution, catalog-specific practice differs from blog-image practice in three ways:
- Bulk compression, not one-off edits. A catalog with hundreds of SKUs needs a compression workflow (batch WebP conversion at upload) rather than manual editing per image. Camera-original files routinely exceed 5MB each, well past what any page needs to load quickly.
- Filenames should carry the part or product identifier, not just a generic description:
stainless-flange-2in-npt-florida.jpgoutperformsproduct-photo-14.jpgfor both search relevance and internal asset management. - Product structured data belongs on catalog pages, not just blog content, since it’s what allows a specific SKU to surface with price and availability directly in image and shopping results.
Frequently Asked Questions
Does image file size affect Google rankings? Yes. Large images slow page load time, and Largest Contentful Paint is a confirmed Core Web Vitals ranking signal.
What is the best image format for SEO? WebP, for most general web use. It offers stronger compression than JPEG or PNG without a visible quality loss, and it’s Google’s recommended format.
Does FAQ schema still matter if the rich result is gone? Yes, but for a different reason than before. As of May 7, 2026, per Google’s own FAQPage documentation, FAQ content no longer produces the visual dropdown in Google Search results, ending a narrowing process that started in August 2023. It’s still read by AI Mode as part of how Gemini verifies and cites content during answer synthesis, so implementing it correctly remains worthwhile, the payoff has shifted from a search-results feature to an AI-citation signal.
How does Google Lens affect SEO? Google Lens matches a photographed object against indexed images using the image itself, not typed keywords. Clear, high-resolution images with descriptive filenames and alt text are more likely to be matched correctly.
Where to Start
Local SEO and technical site performance both depend on how well a site’s images are optimized. Digital Footprint covers the broader technical and structured-data foundation this workflow builds on.
