Clear structure creates useful routes through the information we share.
Clear headings and thoughtful structure help more people use the information we share
By Arbab Naseebullah Kasi, Chief Executive Officer - CEO, Feel Worldwide Foundation Inc.
Imagine receiving a training guide with the answer to your question somewhere inside it. The pages look polished. There are coloured panels, photographs and large section titles. You know the information is there, but finding the part you need means working through the document from the beginning.
For a person using a screen reader, a heading that only looks like a heading may offer no useful shortcut. A beautifully arranged page may also have an underlying reading order that sends the reader from the middle of one explanation into another.
This is an illustrative situation, but the design problem is practical. We often review a document by looking at it. We should also ask how somebody will move through it, locate an answer and understand the relationships between its parts.
That question belongs in the first draft. Whether we are preparing learning materials, a community notice or a small organisation's annual report, accessibility needs a place in the ordinary process of making the document.
Give headings a job beyond appearance
A section heading has at least two responsibilities. Its wording should help the reader recognise what comes next. Its structure should allow software to recognise that a new section has begun.
In a word processor, making a line larger and bold changes its appearance. Applying a built-in heading style identifies its role. Microsoft recommends descriptive headings, built-in styles and a logical hierarchy that does not skip heading levels. Its guidance also distinguishes the document's Title and Subtitle styles from section headings. Microsoft's guidance on accessible Word documents.
For a volunteer handbook, headings such as “Preparing for your first shift” and “Who to contact when plans change” tell readers more than “Introduction” and “Other information.” A reader looking for help with a cancelled shift should not have to guess which broad label conceals the answer.
The structure matters after export too. W3C explains that properly marked headings in a PDF can let assistive technology users access a list of headings and jump to a relevant section. The technical label behind a heading therefore supports a familiar reading habit: looking over the contents before deciding where to begin. W3C's PDF heading technique.
A useful editing exercise is to read only the headings. Do they describe the actual contents? Does the sequence make sense? Could someone tell where a particular question would be answered? Repairing that outline often improves the writing as well as the navigation.
Check the order beneath the page
A page can offer several possible routes to the eye. There may be two columns, a quotation in the middle, a box at the side and a caption beneath a photograph. The reader needs a meaningful route through all of them.
W3C's PDF guidance explains that reading order depends primarily on the order of tagged elements, with content inside those tags also affected by the PDF's content structure. A poorly tagged two-column page may be read across both columns rather than down the first and then down the second. W3C recommends establishing the structure in the authoring tool and checking complex layouts after conversion. W3C's guidance on PDF reading order.
Consider an illustrative course guide. One column explains eligibility. Another explains fees. If sentences from the two sections become interleaved, the reader receives fragments that were never intended to be read together. Every individual word could be accurate while the combined message becomes confusing.
Before choosing an elaborate layout, ask what it contributes. A plain sequence of sections may communicate the same information with fewer relationships to repair. Where columns or sidebars genuinely help, leave time to check the exported file rather than assuming that the screen view proves the order is correct.
This is also a reason to involve the person preparing the final file before the design is settled. Structure is easier to discuss while choices are still open.
Keep words available as words
A scanned page may look almost identical to the original, but it can contain only a picture of the writing. Adobe distinguishes that image from searchable text and explains that optical character recognition, usually called OCR, is needed to recover text from scanned images before other accessibility work. Adobe's overview of PDF accessibility.
That distinction matters when an organisation retrieves an old policy or receives a flyer designed as a single image. Someone should not have to read every letter visually to obtain essential information.
Where possible, return to the editable source. If only a scan remains, recognise the text and review it against the original, especially names, dates, amounts and instructions. A mistaken digit in a contact number has a practical consequence even if the page looks unchanged.
Selectable text is a useful starting check, but it does not establish that the whole document is accessible. Structure, reading order and other features still need attention. Treat the recovered text as material to verify, not as a finished repair.
The same care applies to information placed inside pictures. An event's time and venue should be available in ordinary text, rather than existing only inside a promotional graphic. The design can still carry personality without becoming the only place where the details live.
Describe what the reader needs from an image
Alternative text should reflect the purpose of an image in its setting. The question is not simply what objects appear in it, but what the reader needs to understand.
W3C's image decision tree separates informative images, functional images, complex information and decoration. A meaningful photograph may need a brief description. A complex chart needs its information available elsewhere on the page. A purely decorative web image can have an empty alternative text attribute. The exact implementation differs between web pages and document software, but the decision starts with the image's role. W3C's alternative text decision tree.
For example, a photograph of a building might illustrate a story about neighbourhood architecture. In an arrival guide, the same photograph might be there to identify the correct entrance. The useful description changes with the job the image is doing.
Similarly, writing “bar chart” does little for someone who needs the comparison shown in the bars. Put the important values and conclusion into nearby text or an appropriate accessible data presentation. A caption can explain why the chart is included, while the alternative description helps convey its content.
Avoid treating the description field as a place for promotional keywords. Read it in context and ask whether it provides useful information or merely repeats words the reader has already encountered.
Let links and language explain themselves
A reader should be able to make a reasonable choice before opening a link. “Course timetable” is more useful than several separate links all labelled “click here.” Microsoft notes that some screen reader users review a list of links, making meaningful link text especially important. Its guidance also warns against using colour alone to communicate information. Microsoft's accessible document guidance.
In an illustrative programme guide, “Required sessions” and “Optional sessions” are clearer labels than expecting everyone to distinguish blue dates from green dates. Colour can reinforce the distinction, but the words should carry it too.
Language deserves the same attention. Explain an unfamiliar abbreviation when it first appears. Put the action beside the relevant condition. If a reader must bring a document, name it where you explain the appointment, rather than leaving the requirement in a distant note.
An organisation may be comfortable with its own terminology because staff use it every day. A new participant has not had that practice. Asking a colleague unfamiliar with the programme to follow the instructions can expose assumptions that proofreading for spelling will miss.
Choose the format for the task
A downloadable file can be useful for printing or keeping a copy. It should still be a deliberate choice.
For publishing on government websites, GOV.UK recommends HTML wherever possible so that content can work with users' browser settings. It also cautions that PDFs can create difficulties for finding, using and maintaining information. That is guidance written for a particular publishing context, rather than a claim that every organisation must use the same format for every purpose. UK government guidance on publishing accessible documents.
The broader question is useful elsewhere. Does someone need a formal report to download, or do they need to find the venue address quickly? Could the main instructions be available on a properly structured web page, with a checked document offered for people who want a copy?
Offering more than one format also creates a maintenance responsibility. Decide where the authoritative wording lives and how changes will reach each version. A clear page and an outdated attachment can give two readers different instructions.
The right choice depends on the task and audience. File format alone cannot substitute for thoughtful content and testing.
Make checking part of finishing
Automated accessibility checks are useful, but the result needs interpretation. Adobe's Acrobat guidance includes a “Needs Manual Check” status for items its checker cannot assess automatically. An unresolved manual item is work to review, not a clean bill of health. Adobe's guidance on checking PDF accessibility.
For an ordinary document, make room for a short review before sharing it:
Read the heading outline and confirm that it reflects the document's purpose.
Check that essential information is available as text and that image descriptions fit their context.
Inspect links and any instructions that depend on position or colour.
Review the final exported file, including its reading sequence, rather than checking only the editable source.
Run the available checker, resolve its findings and arrange knowledgeable manual review where needed.
These are starting practices, not a complete certification checklist. A complex report, form or high-stakes document may require specialist assessment. People who use assistive technology can provide valuable feedback, but one person's experience should not be treated as proof that every reader's needs have been met.
Within a team, give this work an owner and enough time. If every document must be repaired immediately before publication, improve the source template and the instructions people receive when they begin. Keep a clear way for readers to report a barrier and obtain the information they need while a repair is underway.
We put care into documents because we want people to use them. That care should extend to the route through the information. A reader looking for one answer should have a reasonable chance of finding it without asking someone else to navigate the document on their behalf.
