How to Find and Fix Bad Alt Text in WooCommerce

An image can have alt text and still have an alt-text problem.

These all technically contain text:

image

product

IMG_4821

shoe

best black running shoes buy online

But none of them necessarily gives a useful description of the image.

That distinction matters in WooCommerce.

A store with 5,000 product images may have very few empty alt attributes and still contain thousands of weak, duplicated, irrelevant, or filename-based descriptions.

So a useful image audit should not ask only:

Does this image have alt text?

It should also ask:

Is the alt text actually useful for this image?

Missing alt text and bad alt text are different problems

Missing alt text is easy to detect.

You might find:

<img src="/shoe.jpg">

or:

<img src="/shoe.jpg" alt="">

But this is also a problem:

<img src="/shoe.jpg" alt="image">

The attribute exists.

The description does not tell you anything useful about the image.

The same applies to:

alt="product"

alt="photo"

alt="IMG_4928"

A basic checker may report all of these as complete because the field is not empty.

A useful audit needs to go further.

1. Generic alt text

Generic alt text describes almost nothing.

Examples:

image

photo

product

item

picture

product image

Imagine a WooCommerce product page showing a green ceramic vase.

Alt text such as:

product

technically labels the image.

But it removes almost all of the information contained in it.

A more useful description might be:

Green ceramic vase with ribbed surface

The goal is not maximum detail.

The goal is enough information to understand what the image contributes.

2. Filename-based alt text

Another common WooCommerce problem is alt text copied directly from image filenames.

Examples:

IMG_4821

DSC00282

black-shoe-final

product_4432_main

chair-side-v4

Some filenames contain useful clues.

That does not make them good user-facing descriptions.

For example:

black-leather-wallet-side.jpg

could become:

Side view of black leather wallet

The filename is useful to the content system.

The alt text should be useful to the person consuming the page.

3. Duplicate alt text across different product images

Duplicate alt text is not automatically wrong.

If the same image appears in the same context, the same description may be perfectly reasonable.

The problem starts when different images communicate different information but all receive the same text.

For example, a product gallery contains:

• main image;
• side view;
• interior;
• close-up;
• lifestyle image.

Yet every image says:

Black leather backpack

The descriptions are accurate at a very general level.

But they discard the reason those additional images exist.

A better set might be:

Black leather backpack

Side view of black leather backpack

Interior compartments of black leather backpack

Close-up of leather texture and stitching

Black leather backpack worn over one shoulder

For more on this, see WooCommerce product gallery alt text.

4. Alt text that describes the product but not the image

This is subtler than generic text.

Suppose the product is:

Oak Dining Table

Every image in the gallery uses:

Oak dining table

That may be fine for the main image.

But another image may specifically show:

• the wood grain;
• the leg construction;
• the extension mechanism;
• the table in a dining room.

In those cases, the product name alone loses useful information.

For example:

Close-up of oak dining table wood grain

is more useful for a detail image than repeating:

Oak dining table

Good alt text describes what this particular image adds.

5. Keyword-stuffed alt text

Bad alt text can also contain too much SEO.

For example:

best black leather backpack travel backpack laptop backpack buy black backpack online

This is not a natural description of the image.

It is a list of search terms pretending to be one.

A useful version could simply be:

Black leather backpack with front zipper pocket

Relevant product terminology is fine.

Keyword stuffing is not.

Alt text should make sense if read aloud as a description of the image.

6. Alt text that contains information not visible in the image

Another failure mode is adding product facts that the image itself does not support.

Suppose the image shows a white T-shirt.

Bad:

Organic cotton T-shirt made in Portugal with sustainable dyes

unless those details are actually communicated visually or necessary in context.

The image may support:

White crew-neck T-shirt

Product data can help generate alt text, but it should not become a license to insert every attribute from the database.

A useful rule is:

Do not claim more than the image and reliable context support.

7. Alt text that is unnecessarily long

More descriptive does not automatically mean better.

Consider:

A high-resolution studio photograph showing a black leather handbag positioned at a slight angle against a plain white background with two handles, a gold zipper, a detachable strap, and soft shadows underneath

This may be accurate.

It may also contain a lot of irrelevant photographic detail.

A more useful version might be:

Black leather handbag with gold zipper and detachable strap

Describe the information that matters.

Do not narrate every pixel.

8. Alt text that repeats nearby text unnecessarily

Context matters.

If the product page already clearly identifies the item as:

Luna Oak Dining Chair

the image may not need:

Luna Oak Dining Chair product image of Luna Oak Dining Chair

A description such as:

Side view of oak dining chair with woven seat

may communicate more useful information with less repetition.

Alt text should work with the surrounding page, not duplicate it mechanically.

9. Inaccurate alt text

An existing description may simply be wrong.

For example:

Image:

red handbag

Alt:

Black handbag

Or:

Image:

rear view of shoe

Alt:

Front view of running shoe

This can happen when:

• images are replaced;
• product variations change;
• imports reuse old metadata;
• assets are copied between products;
• bulk scripts generate descriptions from incomplete context.

An inaccurate description is worse than merely imperfect phrasing because it communicates false information.

10. Alt text that should actually be empty

Not every empty alt attribute is a defect.

A genuinely decorative image should normally use:

alt=""

For example, a purely decorative flourish that contributes no information does not need a description.

This is why an audit cannot simply convert every empty alt attribute into generated text.

The real question is whether the image is meaningful.

For ecommerce product images, the answer is often yes.

For decorative interface assets, it may be no.

A good system needs to distinguish those cases rather than maximizing the number of populated fields.

Examples of bad and better alt text

The better descriptions are not better because they contain more words.

They are better because they preserve more of the useful information in the image.

How to tell whether existing alt text should be replaced

Before rewriting an existing description, ask:

Is it accurate?

Does it describe the image that is actually shown?

Is it specific enough?

Would someone understand what the image contains?

Does it distinguish this image where needed?

If several gallery images show different details, does the description preserve those differences?

Is it natural?

Does it sound like an image description rather than a filename or keyword list?

Is it necessary?

If the image is genuinely decorative, an empty alt attribute may already be correct.

If the existing text passes those tests, there may be no reason to replace it.

Do not regenerate every existing alt text

This is particularly important when using AI.

Once generation becomes cheap, it is tempting to rewrite everything.

That is usually the wrong goal.

Suppose an image already has:

Blue ceramic mug with white handle

An AI model proposes:

Blue ceramic coffee mug featuring a contrasting white handle

The new version is not necessarily better.

You have spent compute and introduced change without fixing a real problem.

Existing alt text should normally be replaced when there is a clear defect.

Examples include:

• missing text on a meaningful image;
• generic descriptions;
• filenames;
• inaccurate descriptions;
• keyword stuffing;
• descriptions that fail to distinguish meaningful gallery images.

The objective is not maximum AI output.

The objective is fewer bad image descriptions.

Why simple missing-alt scanners are not enough

A missing-alt scanner can detect:

alt=""

or a missing attribute.

That is useful.

But it will often consider this valid:

alt="product"

And this:

alt="IMG_4981"

And this:

alt="best shoes buy shoes black shoes online"

All three contain text.

All three may still need attention.

For a real WooCommerce audit, it is more useful to classify images into states such as:

• missing;
• decorative;
• useful existing alt;
• generic;
• filename-like;
• duplicate;
• inaccurate;
• suspiciously keyword-stuffed.

That produces a much better basis for deciding what to fix.

How IMSCRIB approaches the problem

IMSCRIB is built around auditing the images on the actual ecommerce site rather than assuming every populated field is correct.

The workflow is:

scan → identify issues → fix → rescan

That matters because stores rarely have one uniform problem.

A catalogue might contain:

• 300 images with good existing descriptions;
• 40 missing alt attributes;
• 120 filename-like descriptions;
• 70 gallery images with duplicated generic text;
• decorative images that correctly use empty alt attributes.

Those groups should not all receive the same action.

The useful job is to find the images where an intervention creates value.

How to fix bad alt text at scale

For a small WooCommerce store, manual editing may be enough.

For a large store, use a structured process.

1. Scan the live site

Start with the images customers actually receive.

2. Classify the problems

Separate missing alt from weak existing alt.

Do not treat every image as broken.

3. Keep useful descriptions

Do not regenerate good alt text unnecessarily.

4. Fix clear defects

Prioritize images with:

• missing descriptions;
• filenames;
• generic placeholders;
• inaccurate text;
• duplicated descriptions that hide meaningful differences.

5. Use product context carefully

Product names, categories, colors, and other metadata can improve generated descriptions, but only when they accurately relate to what is visible.

6. Rescan the storefront

Check that the final values are actually being rendered.

For large catalogues, see adding alt text to WooCommerce images in bulk.

AI is useful for evaluation as well as generation

AI alt-text tools are often presented primarily as generators.

But generation is only half of the problem.

Before generating a replacement, a system can first ask:

• Is the current description already useful?
• Is it generic?
• Does it appear to be a filename?
• Does it describe the image?
• Is it suspiciously stuffed with keywords?
• Does another image use the exact same description despite showing something different?

This evaluation step matters because it reduces unnecessary rewriting.

For more on generation, see AI-generated alt text for WooCommerce.

The ideal system is not one that generates the most alt text.

It is one that changes the right alt text.

Product gallery images need special attention

Product galleries are particularly prone to weak descriptions.

One common pattern is:

Main image:

Black backpack

Side image:

Black backpack

Interior image:

Black backpack

Detail image:

Black backpack

The store technically has complete alt text.

But the gallery has lost most of its descriptive information.

A good audit should recognize that repeated text across distinct gallery images may deserve review.

Variation images can also look correct while being wrong

Variation-heavy WooCommerce stores create another subtle problem.

Imagine three images:

• black shoe;
• white shoe;
• green shoe.

All three use:

Leather sneaker

That description is not completely wrong.

But it misses the visual distinction the variation images exist to communicate.

Better:

Black leather sneaker

White leather sneaker

Green leather sneaker

For more on this scenario, see WooCommerce variation image alt text.

FAQ

What is bad alt text?

Bad alt text is alternative text that exists but does not usefully describe the image. Examples include generic labels such as image, filenames, inaccurate descriptions, keyword stuffing, or duplicated text that hides meaningful differences between images.

Is duplicate alt text bad?

Not automatically. The same image or equivalent images can reasonably use the same description. Duplicate text becomes a problem when different images communicate different information but the descriptions fail to reflect those differences.

Is image bad alt text?

For a meaningful image, usually yes. It identifies the element as an image but provides almost no information about what the image contains.

Should I replace all existing WooCommerce alt text?

No. Keep existing alt text when it is already accurate and useful. Replace text when there is a clear problem rather than rewriting everything simply because automated generation is available.

Are image filenames acceptable alt text?

Usually not. Filenames are primarily asset-management labels. Even when a filename contains useful words, the final alt text should normally be rewritten as a natural description.

How do I audit WooCommerce alt text?

Inspect or scan the images rendered on the live store and classify them by problem type: missing, decorative, useful, generic, filename-like, duplicate, inaccurate, or otherwise problematic. Then fix only the images that need attention.

Conclusion

WooCommerce image alt text is not a binary field.

An image does not become well described merely because its alt attribute contains something.

image

is not the same as:

Black leather Oxford shoe

And:

Black backpack

repeated across six different gallery images is not the same as describing what those six images actually show.

That is why a useful alt-text audit needs to go beyond missing fields.

Find the descriptions that are actually bad.

Keep the ones that already work.

Fix the clear defects.

Then rescan the storefront to make sure the improved descriptions are really live.