Most SEO advice focuses on headings, keywords, and backlinks. Images get a passing mention, "add alt text", and that's it. But Google uses image metadata as a ranking signal, and most websites leave every field except alt text blank.
The Five Metadata Fields
Every image you publish on the web can carry multiple metadata fields. Here's what each one does and why it matters.
1. Filename
The filename is set before upload and becomes part of the image URL. Search engines parse it for content signals.
- Bad:
IMG_20250115_142356.jpg - Good:
freshly-baked-sourdough-bread.webp
Google's own documentation recommends using descriptive filenames. A file called sourdough-bread.webp tells Google what the image contains before it even analyses the pixels.
Tips:
- Use hyphens to separate words (not underscores)
- Keep it descriptive but concise
- Include the primary keyword naturally
- Use lowercase letters only
2. Alt Text
Alt text is the most important metadata field for SEO. It's used by:
- Search engines to understand image content
- Screen readers for accessibility
- Browsers as fallback text when images fail to load
Write alt text that describes what the image shows, not what you want to rank for. Google's algorithms are sophisticated enough to penalise keyword-stuffed alt text.
3. Title
The image title attribute displays as a tooltip when users hover over an image. While it carries less SEO weight than alt text, it provides:
- Additional context for search engines
- User-facing information on hover
- Default naming in CMS platforms like WordPress
4. Caption
Captions are the visible text displayed below an image. They tend to get read more than surrounding body text. People look at images first and naturally read the text directly beneath them.
While captions aren't a direct ranking factor, they:
- Increase time on page (an indirect ranking signal)
- Provide context that improves content comprehension
- Can include natural keyword usage
5. Description
The description field (used by WordPress and other CMS platforms) serves as a longer-form explanation of the image. It's primarily used for:
- Internal search within the media library
- Potential use by some themes and plugins
- Rich snippets in some contexts
How Search Engines Use Image Metadata
Google processes images through multiple signals to determine relevance:
| Signal | Weight | Notes |
|---|---|---|
| Alt text | High | Primary text signal for image content |
| Filename | Medium | Parsed from URL path |
| Surrounding text | High | Content near the image on the page |
| Page title and headings | Medium | Provides topic context |
| Image title | Low–Medium | Secondary text signal |
| Caption | Low–Medium | Visible text near image |
| Image quality and format | Medium | Affects page speed (Core Web Vitals) |
The combination of these signals determines where (and whether) your images appear in Google Images results and how they contribute to your overall page rankings.
The Compound Effect
Each individual metadata field provides a small signal. The real value comes from getting all of them right, consistently, across your entire site.
Consider two blog posts with identical text content:
- Post A: 8 images, all named
image1.jpgthroughimage8.jpg, no alt text, no captions - Post B: 8 images with descriptive filenames, contextual alt text, captions, and titles
Post B sends dramatically more information to search engines about its visual content. Over hundreds of posts, this compounds into a meaningful ranking advantage.
Automating Metadata at Scale
The challenge isn't understanding what good metadata looks like. It's producing it consistently for every image. A single blog post might have 5-10 images. An e-commerce site might have thousands of product photos. A portfolio site might add dozens of images per week.
At that volume, manual metadata entry becomes a bottleneck. Common shortcuts include:
- Skipping alt text entirely (hurts SEO and accessibility)
- Using the same generic text for every image (provides minimal signal)
- Writing metadata for the first few images and rushing the rest
AI-powered metadata generation solves this by analysing each image individually and producing contextual, descriptive text for all five fields. SquishMate does this automatically as part of its compression workflow. Every image gets a descriptive filename, alt text, title, caption, and description before it's uploaded to your site.
Measuring the Impact
After improving your image metadata, monitor these metrics:
- Google Images impressions. Check Google Search Console under the "Search results" tab with the "Image" search type filter
- Image click-through rate. How often users click your images from Google Images
- Page rankings for image-heavy queries. Queries where image packs appear in results
- Core Web Vitals. If you also optimised file sizes, LCP should improve
Give it 4-8 weeks for Google to recrawl and reindex your images before drawing conclusions.
Quick Checklist
- Descriptive, hyphenated filenames on every image
- Unique, descriptive alt text (under 125 characters)
- Meaningful titles (not just the filename repeated)
- Captions on images where editorial context helps
- Descriptions filled in for WordPress media library search
- Context-aware metadata that relates to the surrounding content
- Regular audits of high-traffic pages for missing metadata
Where to Start
Pick your 10 highest-traffic pages in Google Search Console. Check each page's images for missing alt text, generic filenames, and empty captions. Fix those first. That's where improved metadata has the most measurable effect on impressions. Then build the habit into your publishing workflow so new images ship complete from day one.