Image SEO: How to Optimise Alt Text, File Size, and Formats for Google Rankings

Most websites treat images as a design decision. They pick a file, drop it into the page, and move on. The result is often a 3MB JPEG with a filename like IMG_4872.jpg and no alt text — technically functional, but invisible to search engines and a drag on page speed. Image optimization for SEO is the practice of making every image on your site work harder: for rankings, for accessibility, and for the user experience that Google increasingly uses as a ranking signal.

This guide covers what actually matters — not a checklist of obvious tips, but the decisions that separate high-performing pages from ones that stall despite strong content.

If you're looking for expert help in this area, explore how Indexed's technical SEO services can drive measurable results for your business.

What Image Optimization for SEO Actually Means

Image SEO is not simply about adding alt text. It encompasses every technical and contextual decision made around an image: the file format you choose, the size at which it loads, the filename you assign, the structured data you attach, and how the surrounding content frames what the image shows.

Search engines cannot see images the way a person does. Google's image understanding has improved considerably — its Vision AI can identify objects, faces, and scenes with reasonable accuracy — but it still relies heavily on textual signals to understand context. That means the words around an image, the filename, and the alt attribute carry substantial weight in determining whether an image appears in Google Images or contributes to the ranking of the page it sits on.

Why This Matters Beyond Google Images

Most marketers think of image SEO as being about Google Images traffic. That is a narrow view. Poorly optimised images slow down page load times, which affects Core Web Vitals — specifically Largest Contentful Paint (LCP) and Cumulative Layout Shift (CLS). Both are confirmed Google ranking factors. A single uncompressed hero image can tip an otherwise well-optimised page into a poor LCP score, costing it rankings that have nothing to do with content quality.

Choosing the Right File Format

The choice of image format is one of the highest-leverage decisions in image SEO, and it is frequently made by default rather than by design. Most sites default to JPEG or PNG. Both are still widely used, but neither is optimal for modern web performance.

WebP and AVIF: The Modern Defaults

WebP delivers substantially smaller file sizes than JPEG at comparable visual quality. Google's own WebP documentation notes that WebP lossy images are typically 25–34% smaller than comparable JPEG files. AVIF, a newer format, can compress even further — though it is still building browser support. For most production sites in 2024, WebP is the practical default for photographs and complex images; SVG remains the correct choice for logos, icons, and anything vector-based.

If your CMS or hosting stack does not support WebP conversion natively, tools like Cloudinary, ShortPixel, or Imagify handle the conversion and serve the appropriate format based on browser support via the <picture> element.

When PNG Is Still the Right Choice

PNG is lossless. For images that require pixel-perfect transparency — such as product cutouts on white backgrounds or screenshots with fine text — PNG remains appropriate. The mistake is using PNG for photographs, where file sizes become unnecessarily large without meaningful quality benefit.

Compression and Page Speed: The Ranking Connection Most Teams Miss

Google has been explicit that page experience signals — including Core Web Vitals — influence rankings. What many teams miss is how directly image compression connects to those signals, particularly LCP.

LCP measures the time it takes for the largest visible element on a page to render. On most content pages, that element is an image: a hero banner, a featured photo, or a product shot. If that image is uncompressed or poorly sized, the LCP score suffers regardless of how well the rest of the page is built.

Serving Responsively Sized Images

One of the most common and most fixable image SEO problems is serving a 2000px-wide image to a mobile device with a 390px-wide screen. The browser downloads the full file and then scales it down — wasting bandwidth, slowing load time, and penalising your Core Web Vitals score for mobile users.

The fix is the srcset attribute, which tells the browser to choose from multiple image sizes depending on the viewport. Most modern CMSs generate these variants automatically if correctly configured. If yours does not, it is worth addressing as a technical priority rather than a cosmetic one.

Lazy Loading and the LCP Exception

Adding loading="lazy" to images defers off-screen images until the user scrolls to them — a legitimate performance optimisation. However, applying it to the first image on the page (the LCP candidate) is a mistake that Google's PageSpeed Insights flags regularly. The browser needs to fetch the LCP image as early as possible, so it should carry loading="eager" or no loading attribute at all. A fetchpriority="high" attribute on the LCP image further signals the browser to prioritise it.

Writing Alt Text That Works for Both Google and Users

Alt text is where most image SEO advice becomes either too vague ("describe the image") or too mechanical ("include your keyword"). Neither framing is particularly useful in practice.

What Alt Text Actually Does

Alt text serves two functions simultaneously: it provides a text alternative for screen readers (and therefore matters for accessibility compliance, including WCAG 2.1), and it provides Google with a textual description of the image's content and context. These two functions are aligned, not in tension. Alt text written clearly for a screen reader user is almost always the correct alt text for Google too.

The formula that works in practice: describe what the image shows, in plain language, in one sentence. If the image is genuinely relevant to a keyword topic, that keyword will appear naturally. If it does not appear naturally, forcing it creates awkward alt text that neither serves users nor impresses Google.

Decorative Images and Empty Alt Attributes

Not every image needs a description. Decorative images — dividers, background textures, abstract shapes used for visual styling — should carry an empty alt attribute (alt="") rather than a description. This tells screen readers to skip the image entirely, and it signals to Google that the image carries no meaningful content. Failing to do this means screen readers announce decorative images as unlabelled, which creates an accessibility problem, not a benefit.

Filenames, Structured Data, and the Signals Competitors Skip

Most image SEO guides stop at alt text and compression. Two further signals are consistently underused: filenames and image structured data.

Descriptive Filenames

Google explicitly states in its Google Images best practices documentation that the filename of an image provides useful context about the subject matter. black-leather-office-chair.webp is meaningfully better than DSC00394.webp — not because Google cannot understand images at all without filenames, but because every consistent signal reinforces the same interpretation. Use hyphens to separate words, keep filenames concise, and make them descriptive of what the image actually shows.

Image Sitemaps and Structured Data

If your images are embedded in JavaScript-rendered content or are otherwise difficult for Googlebot to discover through standard crawling, an image sitemap extension tells Google exactly where your images are and what pages they belong to. Google's documentation on image sitemaps provides the correct schema for including image metadata within an existing XML sitemap.

For product pages, recipe pages, or any content type with a defined schema, including image properties within your structured data markup gives Google a direct, machine-readable link between the image and the content entity it represents. This is particularly valuable for e-commerce sites where product images are a primary driver of Google Shopping and Image search traffic.

FAQ

What is image optimization for SEO?

Image optimization for SEO is the process of preparing images so they load quickly, are understood correctly by search engines, and contribute positively to a page's rankings. It covers file format selection, compression, alt text, descriptive filenames, responsive sizing, and structured data — all of which affect both page performance and how Google interprets visual content.

Should I include keywords in alt text?

Only if the keyword accurately describes what the image shows. Alt text should describe the image honestly and clearly. If the image is genuinely relevant to a keyword topic, the keyword will appear naturally. Writing alt text purely to insert keywords — particularly on images that do not depict the topic — can be read as keyword stuffing and provides no accessibility value.

Which image format is best for SEO?

WebP is the current practical default for photographs and complex images, offering smaller file sizes than JPEG without meaningful quality loss. SVG is correct for logos and icons. AVIF offers further compression gains but has incomplete tooling support. PNG is appropriate where lossless quality and transparency are required, such as product cutouts or annotated screenshots.

What SEO tools help with image optimization?

Several tools support image SEO workflows. Google's PageSpeed Insights and Lighthouse audit individual URLs for image-related performance issues. Screaming Frog can crawl a site and identify missing alt text, oversized images, and missing image dimensions at scale. For compression and format conversion, Cloudinary, ShortPixel, and Imagify are widely used. Ahrefs and Semrush's site audit features flag image issues as part of broader technical SEO audits.

What to Do This Week

Rather than a general recommendation to "audit your images," here are the specific first steps worth taking in the next five working days:

• Run PageSpeed Insights on your three highest-traffic pages. Look specifically at the LCP element identified in the report. If it is an image, check whether it is WebP format, whether it has a fetchpriority="high" attribute, and whether it is dimensioned correctly for mobile viewports.
• Use Screaming Frog (free up to 500 URLs) to export all images with missing or empty alt text. Prioritise images on pages that currently rank on page two for target keywords — alt text improvements here are low-effort, high-leverage fixes.
• Check your CMS image upload settings. Determine whether it auto-converts uploads to WebP. If not, identify whether a plugin or CDN-level setting can enable this without a development sprint.
• Review filenames on your five most commercially important product or service images. If they contain camera-generated strings or generic names, rename and re-upload them with descriptive, hyphenated filenames before your next content update.
• Confirm your LCP image is not lazy-loaded. Search your page source for the LCP image URL and verify it does not carry loading="lazy". If it does, remove the attribute — this is a one-line fix that can produce a measurable LCP improvement.

None of these steps require a significant development commitment. Most can be completed or at least scoped within a standard working week, and each addresses a specific, measurable signal rather than a vague recommendation to "optimise your images."

Related Reading

• What Are Common SEO Mistakes?
• Why Does SEO Still Matter in 2026?
• Content SEO: How to Create Content That Ranks, Converts and Gets Cited by AI
• What Is Topical Authority and How Do You Build It?
• How Google's AI Overviews Work (And How to Appear in Them)