Product images and alt text

A shopper looking for a navy fleece opens Google Images before they open a single store, because they want to see it before they read about it. Your photo is either in that grid or it isn't, and the decision is made by text around the photo that the shopper never reads. Product images earn their place by loading fast, describing what they show, and being the one image your structured data points Google to.

Google reads the text around a photo

Google's Images best practices document is clear about where its understanding of an image comes from: the filename, the alt attribute, any caption, the text near the image, and structured data. Its guidance leans on what you write around the photo, which is why the same supplier photo can rank on one store and never appear for another.

Three places carry that meaning:

1. The filename. Google's own example prefers my-new-black-kitten.jpg to IMG00023.JPG. For a store that means ridgeline-3-trail-shoe-navy.jpg rather than a camera export code, renamed before upload, because most platforms keep whatever name arrives.
2. The alt text, covered in the next section.
3. The surrounding text. Google's guidance says to place images near relevant text, on pages relevant to the image's subject. A product photo on the product page beside the product's name and description is exactly that; the same photo in a homepage slider with no caption is not.

Write alt text for the shopper who can't see the photo

Alt text exists for people using screen readers and for browsers that fail to load an image; Google reads the same attribute. Write it for the first audience and the second is served.

Describe what the image shows, in a buyer's words, specific to that image. Google's guidance warns against filling the attribute with repeated keywords; its bad example is an alt that lists "puppy dog baby dog pup" and so on, and its good example describes the scene. For a trail shoe, the main shot gets "Ridgeline 3 trail running shoe in navy, side view", the detail shot gets "Ridgeline 3 outsole showing the lugs and heel", and the lifestyle shot gets "Runner wearing Ridgeline 3 shoes on a muddy trail".

A few rules make this quick:

• Every image on the page gets its own alt, not the product name repeated.
• Variant images name the variant, so the navy photo and the red photo read differently.
• Decorative icons get an empty alt attribute so screen readers skip them.
• Leave out "image of" and "photo of"; the screen reader already announces that.

Your platform's image editor has an alt field for each image. Many themes fall back to the product title when that field is blank, which is why every photo on most stores ends up saying the same thing.

The main image decides your load time

The largest thing in the first screen of a product page is almost always the main product photo, and that is the element Core Web Vitals measure as Largest Contentful Paint. web.dev's Core Web Vitals thresholds call a Largest Contentful Paint of 2.5 seconds or under good, measured at the 75th percentile of visits, and Google's page experience documentation says its ranking systems use Core Web Vitals. A camera-sized photo served at full size to a phone is how a store fails that measure on every product page at once.

Four habits fix most of it:

1. Upload images at a sensible size and let the platform generate the display sizes. Google's image guidance recommends responsive images, where the browser picks a smaller file for a phone; modern themes do this, older ones serve one file to everything.
2. Serve modern formats such as WebP or AVIF where the platform supports them; the image services built into the major platforms convert automatically.
3. Lazy-load images below the first screen, and never the main product image. Deferring the hero delays the exact element the metric measures.
4. Keep the zoom image as a separate larger file that loads only when the shopper zooms.

Store speed and themes covers how to measure this, and Core Web Vitals explained covers the metrics themselves.

Getting into image and shopping results

Google's Images guidance says structured data can put your images into rich results, and its merchant listing documentation makes an image a required property of Product markup. That image is the one Google shows beside your price in results, so it has to be the product. The same documentation recommends several high-resolution images per product, each at least fifty thousand pixels when width is multiplied by height, in 16:9, 4:3 or 1:1 aspect ratios, and it requires the image to represent the product being marked up. A logo or a placeholder in that field forfeits the feature.

Two more things help the photos get found. Google's guidance recommends an image sitemap to help discovery, useful when your images live on a CDN domain. And the crawler has to be able to fetch the image file; an image host blocked in robots.txt can't be indexed.

What to take away

• Google understands a product photo from its filename, alt text, nearby text and structured data, so the same photo can rank on one store and not on another.
• Alt text describes what each image shows for a shopper who can't see it, one description per image, with no repeated keywords.
• The main product image is usually the Largest Contentful Paint element; size it for the screen, serve a modern format, and never lazy-load it.
• Product markup requires an image, and Google recommends several high-resolution photos per product that show the product being marked up.

Next

One product, six colours, four sizes: now decide how many pages that should be. Continue with Variants and duplicate content.