PDF Remediation for Canvas Courses: The Complete Guide

If your Ally score is stuck in the 40s or 50s and you can’t figure out why, there’s a good chance untagged PDFs are the reason. A single untagged PDF scores 0% in Ally’s accessibility check, and that zero gets averaged into your overall course score just like everything else. Upload ten lecture slide decks as “Print to PDF” exports and your course average craters — even if every other piece of content is perfect.

This guide explains what PDF remediation actually means, what the relevant standards are, and your realistic options for fixing the problem. Whether you have five PDFs or five hundred, at least one of these approaches will work for your situation.

What is a “tagged” PDF?

Most people think of a PDF as a static image of a page. And visually, that’s basically what it is. But underneath the visible layer, a PDF can contain invisible structural markup — a set of tags that describe what each piece of content actually is. Headings, paragraphs, figures, tables, lists, and reading order are all encoded in this tag structure.

Screen readers rely on this tag tree to navigate a document. Without it, a screen reader encounters the PDF as a flat, unstructured wall of text. It has no way to distinguish a heading from body text, no way to identify figures, no way to navigate between sections. For a blind student trying to use your lecture slides, an untagged PDF is essentially unusable.

The problem is that most PDFs in the wild are untagged. When you use “Print to PDF” from your browser, or “Save as PDF” from PowerPoint without the right settings, or export from most design tools, the resulting file contains no structure tags at all. It looks fine on screen. It’s completely inaccessible to assistive technology.

What is PDF/UA?

PDF/UA (Universal Accessibility) is the international standard for accessible PDF documents, formally known as ISO 14289. It defines a set of requirements that a PDF must meet to be considered accessible. Ally checks against a subset of these requirements when scoring your files.

The key requirements of PDF/UA include:

• Every content element must be tagged. Text, images, tables, lists — nothing should exist outside the tag tree.
• Reading order must be defined. The tag tree must specify the logical order in which content should be read, which may differ from the visual layout.
• Images must have alternative text. Every figure needs a text description, or must be explicitly marked as decorative.
• Tables must have defined headers. Data tables need header cells (<TH>) so screen readers can associate data with its column or row label.
• The document must have a title. The metadata title field must be set (not the filename — the internal document title).
• The document language must be specified. The PDF must declare its primary language so screen readers use the correct pronunciation engine.

You can check whether a PDF meets PDF/UA requirements using the free PAC (PDF Accessibility Checker) tool, which is the most widely recognized validator for PDF/UA compliance. Ally uses similar (though not identical) checks internally.

Why Ally cares so much about PDF tags

Ally’s scoring model treats PDF tagging as essentially pass/fail. An untagged PDF scores 0%. There’s no partial credit for a PDF that has some tags but not others, or one that’s “mostly” accessible. From Ally’s perspective, if the tag tree is missing, the document is completely inaccessible — which, to be fair, it is.

This binary scoring creates a dramatic effect on your course average. Imagine a course with 20 content items. Nineteen of them score 100%. One untagged PDF scores 0%. Your course average drops from 100% to 95%. Now imagine the more typical scenario: you have 15 tagged HTML pages scoring well, and 10 untagged PDF lecture decks. Those ten zeros drag your average from 100% down to roughly 60%.

This is why PDF remediation is the single highest-leverage thing you can do for your Ally score. Fix the PDFs first and the score moves more than any other single change.

Your options for fixing PDFs

There are four realistic approaches to PDF remediation. Each has different tradeoffs in terms of cost, time, and quality. Here they are, starting with the most manual.

Option A: Adobe Acrobat Pro — manual tagging

Adobe Acrobat Pro includes a full set of accessibility tools for manually tagging PDFs. The process looks like this: open the PDF, navigate to the Accessibility tools panel, run the “Make Accessible” wizard, and then manually review and fix every tag in the document. You’ll need to verify the reading order, add alt text to every image, mark table headers, set the document title, and declare the language.

For a simple 10-page document with straightforward text and a few images, this takes 30 to 60 minutes. For a complex document with tables, multi-column layouts, or dozens of figures, it can take several hours. Multiply that by the number of PDFs in your course and you can see why most instructors never finish the job.

Cost: an Acrobat Pro subscription (around $23/month). The real cost is your time.

Option B: Vendor remediation services

Several companies offer professional PDF remediation as a service. You send them your PDFs, their specialists manually tag them, and they send back compliant files. The quality is typically excellent — these are trained accessibility professionals doing the work.

The downsides are cost and turnaround. Professional remediation typically runs $200 to $600 per document, depending on page count and complexity. Turnaround time ranges from a few days to several weeks during busy periods (like the start of a semester, when every university is trying to hit compliance deadlines simultaneously).

For a handful of critical documents, this is a perfectly reasonable option. For 50 or 100 PDFs across multiple courses, the math stops working.

Option C: Recreate from the source file

If you still have the original Word document or PowerPoint file that you exported to PDF, you can sometimes fix the accessibility issues in the source and re-export. Microsoft Word and PowerPoint both have built-in accessibility checkers and can produce tagged PDFs — if you configure the export settings correctly.

This approach only works under specific conditions: you must have the original source file, the source file must be in a format that supports accessibility features (Word, PowerPoint, InDesign), and you need to know the correct export settings. In Word, for instance, you need to use “Save As > PDF” with the “Document structure tags for accessibility” checkbox enabled. Most people don’t know this option exists.

The other challenge is that many PDFs in a typical Canvas course weren’t created by the instructor. They’re publisher materials, departmental documents, or files inherited from a previous semester. No source file means this option is off the table.

Option D: FixAllyScore — automated PDF/UA tagging

FixAllyScore takes a different approach: automated structural tagging. You point it at a PDF (or a folder of PDFs), it analyzes the content, and it produces a tagged version that passes accessibility checks. The tool runs locally on your machine — your files never leave your computer.

Processing time is measured in minutes, not hours. A typical 30-page lecture deck takes under a minute. A batch of 50 PDFs might take 10 to 15 minutes.

Automated tagging won’t match the precision of a human specialist working in Acrobat Pro for hours. But for the vast majority of course materials — lecture slides, reading summaries, syllabus documents, assignment handouts — the output is accurate enough to move your Ally score from 0% to the 80–100% range on each file.

What FixAllyScore does to your PDFs

When FixAllyScore processes a PDF, it performs several operations to build a compliant tag structure:

• Heading detection. The tool analyzes font size, weight, and spacing to identify headings and assign the correct heading levels (H1, H2, H3). This gives screen readers the ability to navigate by section.
• Figure identification and alt text. Images are detected and tagged as figures. Where possible, the tool generates descriptive alt text. For simple charts and diagrams, this works well. Complex or ambiguous images may need manual review.
• Table structure. Tables are identified and tagged with proper row and column structures, including header cell designation. This allows screen readers to announce which column a data cell belongs to.
• Paragraph and list tagging. Body text is tagged as paragraphs. Bulleted and numbered lists are tagged with proper list structures so screen readers announce “list of 5 items” rather than reading bullets as random characters.
• Document metadata. The document title is set from the filename or detected title text. The language attribute is set to English (or the appropriate language). These are two checks that Ally flags frequently and that are trivial to fix automatically.
• Tag tree construction. All of the above is assembled into a proper tag tree that defines reading order and document structure. The resulting file is designed to pass PAC (PDF Accessibility Checker) validation.

What automated tagging can’t fix

No automated tool handles every edge case. It’s important to understand the limitations so you know when manual intervention is still necessary:

• Scanned image-only PDFs. If a PDF is just a scan of a printed page — a photograph of text rather than actual text characters — no tagging tool can help until you run OCR (optical character recognition) first. FixAllyScore will flag these files so you know which ones need OCR before they can be tagged.
• Complex mathematical notation. Equations, formulas, and mathematical expressions require specialized MathML markup for full accessibility. Automated tagging can identify these regions but can’t generate correct MathML. If your course is math-heavy, the equations will still need attention.
• Hand-drawn diagrams and ambiguous images. Alt text generation works well for standard charts, photos, and simple diagrams. But a hand-drawn flowchart, an architectural sketch, or a highly specialized scientific figure will need manually written alt text that conveys the pedagogical intent of the image.

For most courses, these edge cases represent a small fraction of total content. The automated approach handles the bulk of the work, and you can spend your limited time on the files that genuinely need human judgment.

Start with your PDFs

If you take away one thing from this guide, let it be this: PDFs are the biggest score lever you have. They’re the most common source of 0% scores in Canvas courses, and fixing them produces the largest per-item score improvement. A single remediated PDF can move your course average more than fixing a dozen other minor issues combined.

Start by checking your Canvas course for untagged PDFs. Look at the Ally indicators — the files with red gauges are almost certainly untagged. Pick the approach that fits your situation: manual tagging if you have a few files and the time, source file re-export if you have the originals, or automated tagging if you need to process a batch quickly.

Whatever you choose, start with the PDFs. Everything else is optimization.