Optimize images for SEO: 2026 Alt Text + Filenames

Alt-Text and Filenames: The 2026 Guide to Optimize Images for Search

Introduction: Alt text + filenames are still the fastest way to optimize images for SEO in 2026

If you’ve ever audited a site and found yourself staring at a media library filled with files named IMG_4920.jpg or final-final-v2.png, you know exactly where image SEO goes wrong. It’s usually an afterthought—something fixed hastily right before hitting publish, or worse, ignored entirely until a technical audit flags thousands of errors.

The problem is that without clear text signals, your images are essentially invisible to search engines. Even with the advancements in AI vision we see in 2026, Google still relies heavily on the text you provide to understand context, relevance, and intent. If you want to optimize images for SEO effectively, you need a system, not just a hope that the algorithm figures it out.

In this guide, I’m going to walk you through the exact process I use to turn image optimization from a messy chore into a repeatable workflow. We’ll cover accurate alt text, clean filenames, modern formats like AVIF, and the specific “do’s and don’ts” that keep your site fast and accessible.

How search engines (and generative search) understand images in 2026

Here is the reality of how search engines work today: if a screen reader can’t understand your image, there is a high probability Google can’t understand it either. While computer vision has improved, search crawlers are still fundamentally text-hungry machines.

Search engines index images based on a combination of signals. They prefer standard HTML image elements (<img>) over CSS background images because standard tags allow for attributes like alt text and title tags. If I switch a hero image from a div background to a proper image tag, it suddenly becomes indexable content rather than just “decoration.”

We also need to talk about the shift toward multimodal SEO. In 2026, generative search engines (like Google’s AI Overviews or SearchGPT) use images as part of the answer, not just as a separate tab. They look for entity-based connections. Your alt text helps these systems link the visual data (the pixels of a coffee maker) to the semantic concept (the topic of “best home brewing methods”).

The 3 buckets of image SEO signals: on-image text, on-page context, and delivery/performance

When I teach this to new team members, I break it down into three buckets to keep it manageable. Think of it like shipping a product: you need the right label (text), the right shelf location (context), and a fast delivery truck (performance).

• On-Image Text: This is your direct input—alt text and filenames. This is the strongest signal you control.
• On-Page Context: This includes captions, the heading immediately above the image, and the surrounding paragraph text. Search engines use this to verify if the image is actually relevant to the topic.
• Delivery/Performance: This is how the image loads—file format (WebP/AVIF), responsiveness (srcset), and file size. If your image takes three seconds to load, users bounce, and Google notices.

Accessibility isn’t optional: why SEO and WCAG/ADA overlap

I cannot stress this enough: optimizing for accessibility is optimizing for SEO. The guidelines provided by the Web Content Accessibility Guidelines (WCAG) and ADA compliance standards align almost perfectly with what search engines want: clarity and structure.

For us, this means a simple rule: if an image provides information (like a chart or a product photo), it needs descriptive text. If it is purely decorative (like a swoosh or a generic stock photo of a laptop used for spacing), it should be hidden from screen readers. This isn’t just about avoiding lawsuits—though that is a real business risk—it’s about ensuring that everyone, including the bots crawling your site, understands exactly what is on the page.

Alt text in 2026: how I write descriptions that optimize images for SEO (and help users)

Writing alt text is often where writers get stuck. They either overthink it and write a paragraph, or they panic and stuff keywords where they don’t belong. My goal is always clarity within constraints. I aim for descriptions that are descriptive, specific, and concise—ideally between 80 and 125 characters.

I use a simple mental formula when drafting alt text:

[Subject] + [Key Attribute] + [Context/Purpose]

For example, instead of just “shoes,” I write “Red running shoes with foam soles on a track.” This describes the image accurately for a visually impaired user while naturally including the keywords that matter. You don’t need to say “image of” or “picture of”—the screen reader already announces that it is an image. That’s just wasted character count.

Alt text checklist (beginner-friendly)

Here is the checklist I keep pinned for my team. If it passes these points, it’s good to go:

• Be accurate: Does it actually describe what’s in the image?
• Keep it concise: Aim for under 125 characters so screen readers don’t cut it off.
• No redundancy: Never start with “image of” or “picture of.”
• Context matters: Describe the information the image conveys, not just the aesthetics.
• No stuffing: Only use a keyword if it naturally describes the visual content.

When images can skip alt text (decorative vs. informational)

Not every image needs a description. If you have a divider line, a background pattern, or a generic stock photo that adds zero context to the article, you should use an empty alt attribute (alt="").

This tells screen readers, “Skip this, it’s just decoration.” If you leave the attribute out entirely (no alt tag at all), some screen readers will read the filename instead, which is a terrible user experience. A quick test I use: If I removed this image, would the page lose any meaning? If the answer is no, it’s likely decorative.

Keyword use in alt text: how to do it naturally (and when not to)

Can you use keywords in alt text? Yes. Should you force them in? Absolutely not. I view keywords in alt text as a tie-breaker. If I have a photo of a “CRM dashboard showing sales pipeline,” that is a perfectly natural way to include the keyword “CRM dashboard.”

However, if I’m writing about “best CRM software” and I tag a picture of a smiling team member as “Best CRM software for small business team,” that’s manipulation. It doesn’t describe the image, and it creates a disconnect between what a user (or bot) sees and what the text claims. Google is smart enough to spot that misalignment in 2026.

Examples table: strong vs. weak alt text (with business use cases)

Image Type	Weak Alt Text (Avoid)	Improved Alt Text (Use)	Why it’s better
Product Image	“coffee press” OR “IMG_402.jpg”	“Matte black 32oz French press sitting on a wooden kitchen counter”	Specific details (color, size, material) match user intent for specific products.
Team Headshot	“Sarah smiling”	“Headshot of Sarah Jones, VP of Marketing, against a grey background”	Identifies the person and their role, which aids entity search.
UI Screenshot	“screen shot of error”	“Login screen showing ‘Incorrect Password’ error message in red text”	Describes the specific state and text content of the interface.
Data Chart	“Chart of sales”	“Bar chart showing Q3 revenue growth of 15% across email and organic channels”	Summarizes the data insight, which is the actual value of the image.
Decorative Icon	“icon of a checkmark”	alt="" (Empty attribute)	The icon is purely visual; describing it interrupts the reading flow.

Image filenames and folders: simple rules that help Google before it even “sees” the image

Filenames are the first signal a search engine encounters. Before Google even crawls the image content, it sees the URL. If a human can look at the filename and guess what the image is, you are doing it right. If I see coffee-brewing-guide.jpg, I know exactly what to expect. If I see DSC_8831.jpg, I know nothing.

I stick to a few rigid rules for filenames to keep things organized and optimized:

• Use descriptive words: Describe the visual content accurately.
• Use hyphens, not underscores: Google treats hyphens as space separators. It reads seo-guide as “seo guide.” It often reads seo_guide as “seoguide.”
• Lowercase only: Servers can be picky about case sensitivity. Image.jpg and image.jpg can be treated as different files, which causes headaches.
• Avoid stop words: You don’t need “the,” “and,” or “in.” Keep it punchy.

Filename formulas you can copy (products, blog diagrams, screenshots)

I rely on templates so I don’t have to think about naming every single file. Here are the patterns I use:

• Products: brand-product-attribute-color.jpg (e.g., acme-french-press-steel-black.jpg)
• Blog Diagrams: topic-concept-diagram.png (e.g., image-seo-signal-chart.png)
• Screenshots: platform-feature-step.webp (e.g., wordpress-media-library-settings.webp)

Renaming images safely: what to check before you change filenames

Here is a quick warning: do not go back and rename thousands of old images without a plan. Changing a filename changes the image URL. If that image is ranked in Google Images or linked to by another website, renaming it breaks that link.

If I’m auditing a site, I only rename images if:

1. The page has very low traffic and I’m re-optimizing the whole thing.
2. I can implement 301 redirects from the old image URL to the new one (which is often more technical work than it’s worth for minor gains).

For most intermediate marketers, the best strategy is: optimize everything new going forward, and only fix the old stuff if you are doing a major content update.

Formats, responsiveness, and loading: the technical layer of optimizing images for SEO

You can write the best alt text in the world, but if your image is a 5MB PNG file, your SEO will suffer because your PageSpeed score will tank. The technical delivery of your images is a massive ranking factor because it directly impacts User Experience (UX) and Core Web Vitals.

Modern formats like WebP and AVIF are superior to older formats like JPEG and PNG. They offer significantly better compression without losing quality. I almost exclusively use WebP for web content now—it’s supported by every modern browser and keeps files light.

Table: JPEG vs PNG vs WebP vs AVIF vs SVG (what I use and why)

Format	Best Use Case	My Take
JPEG	Complex photos	Reliable classic, but WebP usually beats it on file size.
PNG	Screenshots, images needing transparency	Heavy. Only use if you need crisp text or transparency.
WebP	Photos & graphics (General use)	My default. Great compression, widely supported.
AVIF	Photos & graphics (High performance)	The future. Even smaller than WebP, but check browser support.
SVG	Logos, icons, simple illustrations	Infinite scalability. Essential for sharp logos on mobile.

Responsive images: when to use srcset/sizes and when not to overthink it

Responsive images solve a simple problem: you don’t want to load a desktop-sized image on a mobile phone. The srcset attribute allows the browser to choose the right image size based on the user’s screen.

If you are on a modern CMS like WordPress, this is often handled automatically. But if you have a custom site, ask your developer: “Are we using srcset to serve smaller images to mobile users?” It’s a standard requirement in 2026. Don’t over-engineer it—just ensure you aren’t serving 4000px wide images to an iPhone.

Lazy-loading done right (without hurting UX or indexing)

Lazy-loading (loading="lazy") defers the loading of images until the user scrolls near them. This saves data and speeds up initial page load. However, there is one major exception: never lazy-load your above-the-fold hero image.

If the first image the user sees is lazy-loaded, the browser waits to fetch it, causing a layout shift (CLS) and a perceived delay (LCP). My rule of thumb: lazy-load everything below the first screen fold. Eager-load the hero image.

My step-by-step workflow to optimize images for SEO (alt text + filenames + performance)

Theory is great, but operations are what actually get things done. When I’m publishing a new piece of content—let’s say, a guide on “How to Use a French Press”—I follow a strict workflow to ensure I don’t miss anything. This is the exact sequence I use.

Step 1: Choose the right image (originality, relevance, and on-page context)

First, I select an image that actually adds value. If I’m explaining the “plunge” step, I need a photo of the hand pushing the plunger, not a generic bag of beans. Original photos always outperform stock photos in search because they are unique entities.

Step 2: Prepare the file before upload (format, dimensions, compression, filename)

Before I even log into the CMS, I prep the file on my computer:

• Resize: I scale it down to the maximum display width (e.g., 1200px).
• Format: I export it as a WebP.
• Rename: I change DCIM_102.jpg to french-press-plunging-technique.webp.

Scaling content requires consistency. When we need to scale this workflow across 50 posts a month, I use an AI article generator to standardize the briefs, ensuring that the requirements for image assets are clearly defined before a writer even starts drafting.

Step 3: Upload and write alt text (accessibility-first)

I upload the file. Immediately—before I forget—I fill in the alt text field. For my French press image, I write: “Hand pressing down the plunger on a stainless steel French press coffee maker.” It’s accurate, contains the keyword “French press,” and describes the action.

Step 4: Make it responsive and crawlable (<img>, <picture>, srcset, lazy-loading)

In the editor, I ensure the image isn’t set to “full size” if it’s just a small diagram. I check that my CMS is applying the standard srcset attributes. Since this image is in the middle of the article, I verify that lazy-loading is active.

Step 5: QA and monitor (accessibility checks + performance + image search visibility)

After hitting publish, I do a quick sanity check. I look at the page source to ensure the alt text is there. Once a month, I peek at Search Console’s “Performance” tab and filter by “Image” to see if our visuals are actually gaining impressions. It’s a quick feedback loop that tells me if my filenames and alt tags are working.

Common mistakes (and quick fixes) when you optimize images for SEO

Even pros make mistakes. Here are the most common ones I catch in audits, and how to fix them quickly.

1. Keyword Stuffing in Alt Text:
   
   Mistake: “French press coffee maker best coffee coffee brewing.”
   
   Fix: Rewrite it to describe the image. “Barista pouring coffee from a French press.”
2. Using “Image of…”
   
   Mistake: “Image of a dog.”
   
   Fix: Just “Golden retriever catching a frisbee.” Save the characters.
3. Decorative Images with Alt Text:
   
   Mistake: Describing a background swoosh as “blue curved line graphic design.”
   
   Fix: Set alt="". Clear the clutter for screen readers.
4. Missing Width/Height Attributes:
   
   Mistake: The page jumps around as images load (CLS).
   
   Fix: Ensure your HTML <img> tag has defined width and height attributes so the browser reserves the space.
5. Lazy-loading the Hero:
   
   Mistake: The main banner takes 2 seconds to appear.
   
   Fix: Disable lazy-loading for the first image on the page.

Mistake-to-fix checklist (copy/paste for your team)

• [ ] Filenames: Are they lowercase with hyphens?
• [ ] Alt Text: Is it under 125 chars and descriptive?
• [ ] Format: Is it WebP or AVIF?
• [ ] Size: Is the file size under 100kb (where possible)?
• [ ] Hero: Is the top image NOT lazy-loaded?

FAQs: alt text, filenames, and modern image SEO (2026)

How long should alt text be?

Aim for 80–125 characters. This is the sweet spot for screen readers and search crawlers. It forces you to be concise without losing meaning.

Should I use keywords in alt text or filenames?

Yes, but only if they naturally describe the image. Filenames are a great place for descriptive keywords (e.g., seo-audit-checklist.pdf), but avoid stuffing lists of keywords into alt text where they don’t belong.

What types of images can skip alt text?

Purely decorative images. If the image is just there for visual flair—like a border, a shadow, or a generic background texture—use an empty alt attribute (alt="").

Do modern formats and responsiveness matter for image SEO?

They matter indirectly but significantly. Google rewards fast-loading pages and good User Experience (Core Web Vitals). Using WebP and responsive sizing makes your page faster, which helps your rankings.

Is automated alt-text generation worthwhile?

At scale, yes. If you have 10,000 eCommerce products, AI-generated alt text can be a lifesaver. However, it requires human QA. I recommend using automation for the first draft, then spot-checking a sample to ensure it’s accurate and compliant.

Summary and next actions: what I’d do this week to improve image search visibility

Optimizing images doesn’t have to be a dark art. It’s about consistency, accessibility, and technical hygiene. If you follow the rules of clarity and performance, you’re already ahead of 90% of the web.

Recap:

• Write descriptive, concise alt text (accessibility first).
• Use hyphenated, readable filenames.
• Serve modern formats (WebP) and stop lazy-loading your hero images.

Your Next Actions:

1. Audit your top 10 pages: Check the alt text on your highest-traffic posts today.
2. Fix your template: Ensure your blog template isn’t lazy-loading the featured image.
3. Standardize the process: Share the checklist above with your writers and devs.

Consistency is the hardest part of SEO. It’s about building an SEO content generator workflow that relies on intelligence, structured data, and clear guidelines, ensuring every single image you publish is working for you, not against you.