What makes a PDF actually accessible (it's not just adding alt text)

Ask ten people what an "accessible PDF" is and you'll get ten answers, most of them wrong. The most common one — "it's got the text in it, not just a picture" — is necessary but nowhere near sufficient. A PDF can be full of selectable text and still be completely unusable for someone navigating with a screen reader.

The difference comes down to a word that sounds duller than it is: tagging.

A PDF is two documents at once

Think of a PDF as having two layers. There's the visual layer — what you see on screen and in print: positioned text, lines, boxes, the works. Then there's the structure layer — an invisible outline that says this is a heading, this is a paragraph, this is a form field labelled "Full name", read these in this order.

Sighted readers use the visual layer. Screen reader users rely entirely on the structure layer. If it isn't there, assistive technology is left guessing — and it guesses badly.

So what is PDF/UA?

PDF/UA (ISO 14289-1, the "UA" standing for Universal Accessibility) is the international standard for that structure layer. A PDF/UA-compliant file has a proper structure tree: headings in a correct hierarchy, paragraphs marked as paragraphs, images carrying alternative text, decorative elements flagged as artifacts so they're ignored, and form fields with accessible names and a sensible tab order.

It's the PDF equivalent of WCAG for web pages — and for UK organisations it matters, because the Equality Act 2010 and the public sector accessibility regulations don't stop at your website. A form is a form whether it's on a page or in a download.

The bit that usually goes wrong

Here's the honest problem: most tools that export PDFs produce the visual layer beautifully and the structure layer barely or not at all. You get a gorgeous-looking form that's an accessibility brick wall. Retrofitting tags afterwards — in Acrobat Pro, by hand — is slow, fiddly specialist work that most teams don't have time for.

The fix is to build the structure as you design, so the tags come out correct automatically rather than being bolted on later.

1. 1

   Use real headings

   A document title is a heading level 1; sections are level 2. Don't fake headings by just making text big and bold — that's invisible to the structure layer.

2. 2

   Label every field

   Each input needs a clear, human-readable label ("Full legal name", not "Name1") that's exposed to assistive technology, not just printed next to the box.

3. 3

   Set a deliberate reading and tab order

   Top to bottom, logical grouping — so keyboard and screen reader users move through the form the way you'd expect.

4. 4

   Mark decoration as decoration

   Background shapes and lines should be artifacts, so they're skipped rather than announced.

How FormGenius handles it

This is the entire reason FormGenius exists. When you design a form on the canvas, it builds a tagged PDF/UA structure tree behind the scenes: headings keep their hierarchy, form fields export with their labels and tooltips, decorative shapes are marked as artifacts, and the reading order follows your layout. The Accessibility Wizard then checks the lot — heading hierarchy, alt text, contrast, tab order — before you export, so you find problems while you can still fix them in two clicks.

You design the form. The accessibility comes out correct by default. That's the whole idea.