Skip to main content
Home
GrN.dk

Main navigation

  • Articles
  • Cases
  • Services
  • Your Digital Project Manager
  • About Greg Nowak
  • Image Gallery
  • Contact
User account menu
  • Log in

Join my community / free newsletter — sign up here

Breadcrumb

  1. Home

AI alt text: How to tackle your online store’s image backlog

Illustrated infographic summarizing: AI alt text: How to tackle your online store’s image backlog

By Greg Nowak. Last updated 2026-09-08.

If your online store has lots of images without alt text, it is tempting to have AI fill in all the fields at once. The product images are already there, after all. But the text needs to match the purpose the image serves on the page. Otherwise, you could end up with a lot of new descriptions that still need cleaning up.

Could your business use this? nowa.dk sets up AI automation for Danish companies.

Start with one product category. Map the images, generate suggested text in Danish, and have an editor approve it before publication. Then check the results on the pages. This gives you something concrete to assess: how good are the suggestions, and how much work does it take to get them right?

The image’s function determines the text

A product photo, a decorative image and a clickable icon serve different purposes. W3C’s image guidance, updated on 8 April 2026, explains that the text alternative should depend on the image’s content, function and context. Informative images should convey the essential information, decorative images should have an empty text alternative, and functional images should describe the action or destination. See W3C’s guidance on images.

An empty alt field is therefore a reason to take a closer look at the image. It may be entirely correct. Use the table for an initial assessment:

What purpose does the image serve on the page?
How the image is used What should the text alternative convey? How to handle it in the pilot
Product photo with relevant visual information A brief description of the essential details Create an AI draft and have the editor check it in the context of the page.
Purely decorative image No text content: alt="" Record that the empty text alternative is a deliberate choice.
Image that is the only content in a button or link The action or destination Check what the button or link does on the page.
Chart or illustration containing extensive information A comprehensive text alternative Set the image aside for a separate review.
The image categories are based on W3C’s guidance. The workflow is a suggestion for the online store’s pilot.

Consider a hypothetical example: a close-up of a bag’s interior compartments may need a description of the layout. An icon whose sole purpose is to open the shopping cart needs to say what it does. A general description of what is pictured cannot meet both needs.

Map images where customers encounter them

Create a working list of the images in the pilot category and the pages where they are used. Give each use its own row. If the same image appears on both a category page and a product page, you need to be able to assess the two uses separately.

The list should include the media ID, page URL, existing alt text, product name, image role and review status. Assign a responsible editor as well. This lets you see which images still need a decision and who needs to make it.

Statuses might include: awaiting assessment, decorative, draft ready, approved and checked on the page. Record the decision to treat an image as decorative separately. Otherwise, the same empty field can easily reappear as an unresolved task during the next cleanup.

Agree from the outset how you will handle existing text. A manageable starting point for the pilot is missing text and clearly defined corrections. More substantial rewrites should require a separate decision so the editor knows what is being changed.

Give the AI a specific task in Danish

Send relevant information from the page along with the image: the product name, variant, nearby text and assessed image function. Keep product data separate from the editor’s information. It should be possible to see what the suggestion is based on.

For informative product images, an instruction might read: Write a short Danish text alternative for this use. Only include details supported by the image or the accompanying product data. Avoid sales language and flag any uncertainty separately for the editor.

Save the suggested text and review notes in separate fields. Uncertainty about the colour or product variant needs to reach the editor without becoming part of the text alternative presented to the customer. Include materials, dimensions and features only when they are documented and relevant to the description.

Collect a few internally approved examples from the pilot category. These give the editor a concrete reference for assessing new suggestions: which product terms do you use in Danish, and how much detail should be included? Use the same examples when you later change the instructions and need to assess whether the suggestions have improved.

Choose the integration to suit the workflow

In WordPress, the media REST API can read the alt_text field and update a specific media item. This is described in WordPress’s media documentation. An integration can therefore retrieve existing values and write approved changes back.

A ready-made plugin may also be an option. AltText.ai’s WordPress and WooCommerce integration documents batch image processing, the use of product names and editing of generated text. On private installations, the plugin can be configured to upload images directly for analysis instead of retrieving them from a public image URL. Analysis therefore does not necessarily take place locally, even if the test installation is private.

Try an existing plugin if it fits your workflow. If you require approval before overwriting, separate decisions for each use of an image or a specific change log, the pilot should establish whether the plugin can be configured accordingly. Otherwise, an additional layer may be needed.

Through nowa.dk, Greg’s AI automation service for Danish companies, Greg can build a WordPress review workflow with AI integration, editorial approval and an undo option. The scope depends on how you publish today and which requirements the pilot needs to test.

Approve first, then check the page

The editor should be able to see the image, its context on the page, the existing text and the AI suggestion together. From there, it should be possible to approve, edit or send that use of the image back for assessment. Review every suggestion before publication during the pilot. This makes it easier to spot recurring errors while the task is still small.

Save the previous value along with the change, timestamp and approval. Also agree on what the undo function should actually restore: the media field, the text in the page content or both. Test it on a specific change before processing more images.

An update in the media library is not enough to complete the task. AltText.ai’s documentation explains that, by default, changing alt text in the media library does not update the alt text for images in existing posts and pages. The plugin has a separate function for updating the content. It is described in the integration guide.

The pilot should therefore also include checks of the category listing, product page and gallery. Inspect the actual alt attribute on the published pages, and test relevant images and image links with a screen reader. Record the result for each use. A successful API response alone does not tell you whether the customer encounters the right text.

Be precise about what the image review covers

The accessibility requirements apply from 28 June 2025 to e-commerce services within their scope. Exemptions include microenterprises that provide services. Use Sikkerhedsstyrelsen’s (the Danish Safety Technology Authority’s) guidance on e-commerce services to clarify how the requirements apply to your company.

The pilot’s deliverable is to improve and check the selected uses of images. A completed image review does not establish that the entire online store meets all relevant accessibility requirements. That boundary should be clear in the brief from the outset.

Let one product category guide the next step

Agree on what you will measure before the pilot starts: editor time, AI service usage, suggestions approved without edits, recurring errors and discrepancies on the published pages. With this information, you can calculate the total effort involved for each fully checked use of an image.

The results should help you decide where further work is needed. You may need to adjust the instructions, improve the product data or adapt the integration before moving on to the next category. Also assign responsibility for new images to a role within the company. Otherwise, the same cleanup will soon need repeating.

Greg can help draw up a concrete proposal for the first category, the approval process and the checks on the pages. This gives you a clearly scoped task and a basis for assessing how AI automation fits into day-to-day operations.

Related on GrN.dk

  • Fra salgsmøde til CRM: Automatisér opfølgningen uden datarod
  • Har jeres AI-chatbot husket at sige, at den er en bot?
  • Hvad spørger kunderne om? Lad AI læse mønstrene i supporten

Need help with this kind of work?

Talk to Greg about an alt text pilot for your online store Get in touch with Greg.

Sources

  • W3C: Billeder og tekstalternativer
  • WordPress: Medier i REST API
  • AltText.ai: Integration med WordPress og WooCommerce
  • Sikkerhedsstyrelsen: Tilgængelighed for e-handelstjenester
Last modified
2026-09-08

Tags

  • AI-automatisering
  • wordpress
  • webtilgængelighed
  • e-handel

Review Greg on Google

Greg Nowak Google Reviews

 

Written recommendations from Trafik og Veje, Aarhus Municipality (2011) and AgroTech (2010) — read them on LinkedIn.

Illustrated infographic summarizing: AI alt text: How to tackle your online store’s image backlog
AI alt text: How to tackle your online store’s image backlog
2026-09-08

Use AI for your online store’s alt text with a manageable pilot: map the images, generate suggestions in Danish, and check the results in WordPress and WooCommerce.

Illustrated infographic summarizing: From Supplier PDFs to Product Data: Where AI Needs a Second Check
From Supplier PDFs to Product Data: Where AI Needs a Second Check
2026-09-07

Supplier files need more than extraction. Here’s how to check coverage, match SKUs, resolve unclear units and prices, and test product data before a catalogue import.

Illustrated infographic summarizing: Shorter TLS Certificates: Will Your Renewal Setup Keep Up?
Shorter TLS Certificates: Will Your Renewal Setup Keep Up?
2026-09-06

Shorter TLS certificates leave less room for renewal problems. Check domain validation, scheduling, deployment and the certificate your customers actually receive.

Illustrated infographic summarizing: Your AI Image Has Content Credentials. Will Your Website Keep Them?
Your AI Image Has Content Credentials. Will Your Website Keep Them?
2026-09-05

AI image credentials can disappear during routine website processing. Learn how to test your CMS, optimizer, CDN, and publishing workflow end to end.

Illustrated infographic summarizing: What Are Customers Asking? Let AI Find the Patterns in Support Tickets
What Are Customers Asking? Let AI Find the Patterns in Support Tickets
2026-09-04

AI-based ticket analysis can uncover recurring complaints, product defects and gaps in documentation—without the company needing yet another chatbot.

Illustrated infographic summarizing: OpenAI Has Machine Identity Now. Which Jobs Should Lose API Keys?
OpenAI Has Machine Identity Now. Which Jobs Should Lose API Keys?
2026-09-03

OpenAI’s X.509 workload identity can replace API keys for the right workloads. This practical framework helps teams decide where to start safely.

Illustrated infographic summarizing: WordPress 7.1 Exposes AI-Ready Actions. Who Gets to Run Them?
WordPress 7.1 Exposes AI-Ready Actions. Who Gets to Run Them?
2026-09-02

WordPress 7.1 helps AI agents discover and invoke site abilities. Here is how to keep exposure, authentication and permission firmly separate.

Illustrated infographic summarizing: From Sales Meeting to CRM: Automate Follow-Up Without Compromising Data Quality
From Sales Meeting to CRM: Automate Follow-Up Without Compromising Data Quality
2026-09-01

How to use AI for meeting notes and follow-up while fixed rules protect CRM data, customer matching and the sales pipeline from errors and premature changes.

Illustrated infographic summarizing: Your AI Gateway Can Name the User. Decide What That Log Is For
Your AI Gateway Can Name the User. Decide What That Log Is For
2026-08-31

Identity-aware AI Gateway logs can sharpen security and cost control, but only when attribution, access, retention, guardrails, and response are clearly defined.

Illustrated infographic summarizing: Zero Data Retention Is a Workflow Audit, Not a Checkbox
Zero Data Retention Is a Workflow Audit, Not a Checkbox
2026-08-30

Zero Data Retention covers the provider, not every copy in your stack. See how to audit endpoints, logs, storage, deletion and project-level controls.

More articles

Built by AI — available for your business. The daily articles on this site are researched, written and illustrated by an autonomous AI pipeline. At nowa.dk I install the same kind of AI automation in businesses at fixed prices — site in Danish, English version here, and web/marketing agencies have a dedicated page.

RSS feed

Footer

  • All articles
  • Contact

GrN.dk — AI automation, web platforms, web optimization, data handling and logistics.

© 2026 GrN.dk · LinkedIn · Contact · AI automation in Danish: nowa.dk

Behind GrN.dk: Individual Entrepreneur Codecrafter · Tax ID 305669096 · Bakhtrioni St. 22, 0194 Tbilisi, Georgia · official business register