For
PDF accessibility for universities: course materials at scale
Thousands of documents, dozens of authors, one accessibility officer, and a deadline. What triage looks like when you cannot remediate everything.
The shape of the problem
A university’s PDF estate is not one backlog, it is several with different owners: the public website, the VLE, the library’s digitised readings, administrative forms, and research outputs. They have different volumes, different authors and different deadlines, and treating them as one queue is why programmes stall.
Public web content is smallest and most urgent. It is what a complaint will cite.
Course materials are the largest by an order of magnitude and are produced continuously by people who do not report to you.
Forms matter disproportionately. A student who cannot complete an application has been excluded from the institution, not inconvenienced by a document.
What works
Census before remediation. You cannot plan against an unknown quantity. A validator run across everything, exported as JSON, turns “forty thousand documents” into “eleven thousand need work, of which two thousand need a person”.
Fix the template, not the output. Most administrative documents come from a handful of templates. One corrected Word template fixes every future document from it.
Give lecturers the smallest possible ask. Not a training course — one page: use real heading styles, use real lists, add alt text, tick Document structure tags on export. That covers most of what goes wrong.
Prioritise by consequence. Forms first, then materials for current modules, then the archive. Nothing in any regulation prioritises for you, and pretending everything is equally urgent guarantees the wrong things get done.
Where Taggart fits
The free validator makes the census cost nothing, which is what makes it possible to do properly. Its JSON report means triage across thousands of files is a script.
For the documents that need work, the reading-order editor is where the time actually goes on academic content — multi-column papers are the worst case for auto-tagging, and the case where Acrobat’s tooling is slowest. The remediation report gives the accessibility office a record of what was done to each document and who decided the judgement calls, which is what you need eighteen months later when someone asks.
Frequently asked questions
We have 40,000 PDFs. Where do we start?
With a census, not with remediation. Run a validator across everything and export JSON. You will typically find a third are already usable, a third need only mechanical fixes, and a third need real work. That distribution is the plan. Taggart's validator is free and unlimited specifically so this step costs nothing.
Should we train faculty or centralise remediation?
Both, in that order of leverage. A lecturer who applies real heading styles in Word produces an accessible PDF for free, forever. A central team fixing exports is repairing the same document every semester. Training is slower to land and worth more.
What about documents we did not create?
Published articles, publisher textbooks and third-party materials are the hardest part and the least tractable. Ask the publisher first — many have accessible versions they do not advertise. Budget for the ones that do not.
Check your own file. Taggart's validator is free and unlimited — no page cap, no watermark, no account.
Last updated .