Why Must Documents Be Accessible?
Imagine you are beginning a degree. For the vast majority of people, this is an exciting step, accompanied by considerable preparation and anticipation. Now put yourself in the position of a first-year student with a disability. The lecture halls are wheelchair-accessible, and the pathways across campus are well maintained. Yet when it comes to actually using the learning materials – reading lecture notes, working through presentations, understanding examination documents – some students encounter obstacles that are, at first glance, far less obvious. Not because the content is too demanding (setting aside the intellectual challenge that a degree inherently entails), but simply because the documents are not designed in a way that allows everyone to use them. Accessibility is, after all, far more than a ramp or a lift in place of a staircase.
Let us generalise this example: wherever documents carry information intended for everyone – whether in an educational context, the workplace, public authorities, or everyday life – a lack of accessibility potentially leads to exclusion or discrimination. Someone who cannot read a contract independently, who relies on a screen reader that cannot process a poorly structured PDF, or who struggles with poorly laid-out texts due to a cognitive difference, is effectively excluded from information. And, by extension, from participation in processes that directly affect them.
- Why Must Documents Be Accessible?
- A Few Basics – Fundamental Principles of Accessible Documents
- Accessible from the Start – What to Bear in Mind When Creating the Source Document in Word & Co.
- Preparing for Distribution – Exporting as an Accessible PDF
- Comprehensive Guides for Creating Accessible Documents with Word, PowerPoint, and InDesign
- A Brief Conclusion
Who Actually Benefits from Accessible Documents?
Most obviously, people with visual impairments or blindness who rely on screen readers benefit when documents are designed accessibly. Yet, the circle of those affected is far wider: people with motor impairments benefit from clear navigation and linked headings. People who are deaf from birth, who acquire language differently from the norm, benefit from plain language and a clear structure. Neurodivergent individuals – such as those with dyslexia, ADHD, or autism – benefit from visually calm, consistently designed documents and have, beyond that, highly individual needs regarding documents that work well for them.
A principle that is repeatedly confirmed in the field of accessibility applies here: what is necessary for people with impairments is a useful improvement for everyone else. High contrast and clear heading hierarchies, for instance, make documents more readable and comprehensible for all.
And furthermore: accessibility has long since become a legal requirement. The German Accessibility Strengthening Act (Barrierefreiheitsstärkungsgesetz, BFSG), based on the European Accessibility Act (EAA), and the BITV oblige public bodies – and, specifically since 2025, private companies as well – to provide digital content accessibly. For PDF documents, the PDF/UA standard has established itself as the technical reference.
Accessibility as a Mark of Quality Rather Than a Cost Factor
In my work at the Chair of Human-Computer Interaction at TU Dresden, document accessibility was a recurring and important topic. For the working group Services for Disability and Studies (AG SBS), affiliated with the chair, it is moreover a central focus of their work.
Time and again, I encountered – and continue to encounter – the same pattern: accessibility is treated as an afterthought, to be addressed once a document is finished. This is conceptually misguided and economically unwise. Those who retrofit accessibility after the fact pay considerably more – in time, effort, and cost. The more sensible approach: build accessibility in from the outset as a mark of quality.
This is the approach you should adopt from this point forward – and it requires a degree of know-how. What fundamental principles are at play? What must be considered in Word, PowerPoint, and InDesign? What applies when exporting to PDF? This article provides your first point of entry. Comprehensive guides from a project I initiated at TU Dresden in 2020 together with colleagues are available for download free of charge and will provide the necessary depth should you wish to go further.
A Few Basics – Fundamental Principles of Accessible Documents
Before turning to individual applications, it is worth examining the fundamental principles that apply universally to accessible documents – regardless of whether you are working with Microsoft Word, PowerPoint, or Adobe InDesign. The focus here is primarily on those aspects that ensure documents can be perceived using various devices and assistive technologies – that is, particularly for people with visual or motor impairments.
Important: The list that follows is by no means exhaustive of all elements of good accessibility, but it provides a solid foundation and involves comparatively little effort.
Aspect 1: Semantic Structure and Reading Order
A central concept is the semantic structure of a document. This refers to the fact that documents should have a coherent heading hierarchy and that all elements are arranged in a logical reading order. This is important for two reasons: firstly, it enables visual orientation within the document; secondly, it ensures that screen readers – software that reads document content (mostly) aloud – output information in the correct sequence. Figure 1 shows the outline view for documents in Word alongside the navigable document structure in the left-hand panel, allowing the structure to be monitored at all times.
This becomes particularly relevant in multi-column layouts: without a correctly defined reading order, a screen reader may read line by line across both columns, rather than working through the first column in full before moving to the second. Heading levels should also not be skipped – jumping from a first-level heading directly to a third-level heading makes navigation considerably more difficult.
Aspect 2: Alternative Text for Images and Graphic Elements
Images, diagrams, and graphics are inaccessible to people who rely on screen readers unless they carry alternative text. What matters here is not merely whether an alternative text is present, but whether it is genuinely useful. A straightforward image description (such as: “Photo of a man at a desk”) is usually insufficient. Instead, the alternative text should convey the information that the graphic element communicates in terms of content: what should the reader take away from it? Purely decorative elements, on the other hand, should be explicitly marked as such, so that assistive technologies ignore them and they do not cause confusion. Figure 2 shows how to add alternative text to graphic elements in Word and other Microsoft 365 applications.
Aspect 3: Contrast Ratios and Colour Choice
For text to be legible, sufficient contrast between the typeface and the background must be ensured. The same applies to graphic elements in which relevant information must be visually distinguishable – for instance, in information visualisations and charts. A considered colour palette benefits not only people with visual impairments, but ultimately improves legibility for everyone – for example in poor lighting conditions or on lower-quality displays.
Aspect 4: Accessible Tables and Lists
Tables should never be inserted into a document as a mere image, as this would allow only an image description. Instead, the table structure must be recognisable as such: header rows must be marked accordingly and must allow unambiguous assignment to the relevant content. With lists (cf. Figure 3), care must be taken to ensure that individual list items are clearly delineated from one another, and that in nested lists the hierarchical levels remain structurally discernible.
Aspect 5: Typefaces, Font Sizes, and Line Spacing
Ornate or heavily decorative typefaces impede the flow of reading – particularly for people with dyslexia or visual impairments. Font sizes that are too small are difficult to read; line spacing that is too tight leads to unintentional line-skipping when reading longer texts. Clear, sans-serif or well-legible serif typefaces, a minimum font size of 11–12 points, and generous line spacing are straightforward measures with a perceptible impact. In most applications, this requires adjusting the relevant style (cf. Figure 4).
First Takeaway
These five aspects apply across all applications – regardless of what you are working with. It is important that the embedded information – such as alternative texts or heading levels – is preserved when exporting to another format. This depends on the application used and the export settings, and can be verified at the end. One thing, however, is certain: those who take these fundamentals into account from the very beginning of creating a document significantly reduce the effort required for accessibility afterwards – and deliver, from the outset, a result that benefits all readers.
Accessible from the Start – What to Bear in Mind When Creating the Source Document in Word & Co.
The good news: the most widely used document creation applications already come with built-in tools that support accessible design. What matters, is using them consistently.
Microsoft Word
In Word, the single most important rule is: styles over manual formatting. Those who simulate headings by using a larger, bold typeface rather than the built-in heading styles produce a visually similar result, but structurally create a text block that is difficult to navigate. Screen readers cannot use such pseudo-headings for navigation. Heading levels should always follow a logical sequence. Skipping levels – jumping directly from Level 1 to Level 4, for instance – causes confusion (“Hang on, have I missed something, or has the screen reader withheld visible information from me?”) and should be avoided. You have already seen the outline view in Figure 1 above, where you can check the structure in Word. Figure 5 below shows how you can use the styles panel to assign the appropriate style to selected passages of text with ease.
For images, Word allows alternative texts to be added or elements to be explicitly marked as decorative (cf. Figure 2). Microsoft 365 supports this with AI-generated image descriptions. These, however, typically describe only in broad terms what is visible in an image, rather than what information it is intended to convey. They are therefore frequently of limited use and must be revised manually. Worth highlighting on a positive note is the integrated accessibility checker, which swiftly and straightforwardly draws attention to potential problem areas.
Microsoft PowerPoint
PowerPoint offers similar tools to Word, but has its own particular pitfalls. Of particular importance is the use of predefined placeholders for titles, footers, and text areas – those who instead place these elements manually on blank slides inadvertently lose the semantic markup. Only correctly placed titles are interpreted as slide headings.
A frequently overlooked issue is reading order: elements are read aloud in the order in which they were placed on the slide – not in the visually logical order. Anyone who revises a slide across multiple editing sessions risks ending up with a reading order that no longer corresponds to what the eye perceives. This can be checked and corrected in the selection pane, or alternatively in the dedicated reading order view (cf. Figure 6).
Particular caution is warranted with SmartArt elements: what appears visually structured often yields barely usable semantic information upon export. Separate alternative texts are essential here. Figure 7 shows how this is done. It is, in truth, rather counterintuitive that the structured information from which these graphics are generated is disregarded – but once you are aware of this, it can be managed accordingly.
Adobe InDesign
InDesign offers maximum creative freedom, which simultaneously makes accessibility more demanding. Text is placed within frames and visual elements are positioned freely. The logical reading order must therefore be defined explicitly via the Articles panel. Here too, the same rule applies: use paragraph and character styles consistently, set tags correctly, and configure export settings with care. One advantage: as an Adobe product, InDesign offers a particularly reliable PDF export – provided the groundwork has been laid correctly.
Second Takeaway
The most common mistake across all three applications is the same: manual formatting instead of semantic markup. Those who work consistently with styles, correct structural elements, and carefully composed alternative texts from the outset lay the best possible foundation for an accessible document, regardless of the application used. Many of these principles can, incidentally, be applied directly to comparable applications such as LibreOffice Writer, Apple Pages, Keynote, or Affinity Publisher.
Preparing for Distribution – Exporting as an Accessible PDF
PDF is the standard format when documents are to be shared – whether as a digital file or as a print-ready document. It is platform-independent, widely used, and the first choice in most contexts. Yet, this is precisely where accessibility so often breaks down a carefully structured Word document can, upon export, become a PDF in which semantic details are absent – no heading hierarchy, no alternative texts, no navigable links.
The opposite of a carefully exported document is the scanned PDF: a text document is printed (for instance, to be signed by hand), scanned, and saved as a PDF – the result is, in essence, an image, not a text. Assistive technologies cannot read anything from it. Subsequent optical character recognition (OCR) can reconstruct individual characters, but as a rule yields no semantic information: heading levels, alternative texts, and document structure cannot be recovered in this way. Printing and rescanning should therefore remain an absolute last resort.
What to Bear in Mind When Exporting
When exporting from Word, PowerPoint, or InDesign, several points are critical: the resulting PDF should be exported as a tagged PDF – that is, with embedded structural information that makes headings, lists, tables, and other elements machine-readable. Furthermore, the document language and relevant metadata such as title and author should be set correctly, as this information too is relevant for assistive technologies. The foundation for this is the groundwork you laid when creating the document. Often, you can already use built-in tools to check whether key accessibility criteria have been met, as illustrated in Figure 8 using PowerPoint as an example.
The PDF/UA standard (where UA stands for Universal Access – not to be confused with PDF/A, where the A stands for Archive, denoting long-term archiving) defines authoritatively what constitutes an accessible PDF and serves as the technical reference for accessibility auditing.
Checking Accessibility
Before distributing, it is advisable to check the exported PDF. Tried-and-tested tools include the Adobe Acrobat Accessibility Checker and the free PAC: PDF Accessibility Checker. Both quickly provide a helpful overview of structural issues. However, it must be noted: automated tools can check whether alternative text is present, but not whether it is meaningful in context. Manual review remains, for now, irreplaceable.
Third Takeaway
An accessible source document is by far the best foundation for an accessible PDF. Those who attempt to make corrections in the finished PDF face considerable additional effort, often with incomplete results.
Comprehensive Guides for Creating Accessible Documents with Word, PowerPoint, and InDesign
TU Dresden had for some time maintained internal guides for staff, describing how documents in common office applications can be created to meet a minimum standard of accessibility. Over time, however, these materials had become outdated, and a revised edition was long overdue. As part of a project in which I was actively involved, so-called Inklusionsmittel (budget for inclusion-fostering initiatives) were secured for this purpose: funds from the Free State of Saxony that can be deployed specifically for such initiatives.
The result: a comprehensive set of materials encompassing both detailed guides and compact summary versions in flyer format, each available in German and English. Figure 9 gives a first impression of the printed editions of the guides.
Detailed guides are available for:
- Microsoft Word (versions 2016 and 2019)
- Microsoft PowerPoint (versions 2016 and 2019)
- Adobe InDesign
Summary versions in flyer format, presenting the key points at a glance, are likewise available for all three applications, individually and in both languages. They are aimed at anyone who lacks the time or inclination to work through a full guide: a low-threshold offering as a first point of entry.
The materials were distributed in printed form across TU Dresden and are also freely available for download online. They are expressly not intended solely for members of TU Dresden, but for all those with an interest in the subject – which is precisely why I warmly recommend them to you.
A note on currency of the project results: the guides refer to Office 2016 and 2019, as well as older versions of InDesign. The procedures described are, however, largely transferable, and the underlying principles remain unchanged. Only newer features, such as AI-assisted generation of alternative texts, are not yet covered.
I warmly invite you to have a look at the guides, to pass them on, and to put them to active use – as a practical contribution towards greater accessibility in documents.
Downloads
The materials and further information are available for download on the TU Dresden website:
A Brief Conclusion
As you will have gathered, accessible documents are not an optional nice-to-have. On the contrary: they are a legal requirement in many areas and, in every case, an important contribution to society. The central insight of this article can be distilled into a single sentence: accessibility begins in the source document, not in the PDF. Those who attend to the essential elements during the creation of a document save themselves laborious reworking, reduce costs considerably, and deliver from the outset a result that benefits all readers.
What the Future Holds
It will be fascinating to see what contribution artificial intelligence will make to accessibility in the years ahead. Modern AI models are already capable of describing images in considerable detail – and when the textual context of a document is known, the actual message conveyed by a visual element can increasingly be derived from it. The potential for AI-assisted alternative texts is substantial. For now, however, the following still holds: the results can serve as a helpful aid, but they do not replace manual review and revision. It would be unwise to rely on them blindly.
Equally, promising is a growing awareness of neurodiversity: the more accessibility is understood not merely as a concern for people with visual impairments, but as encompassing the needs of neurodivergent individuals as well, the more inclusive documents – and digital communication as a whole – will become.
