Preserve the design
For most documents, Gevell PDF converts your PDF into an accessible layout — restructured, reflowable, and rebuilt from the content up. But for flyers, posters, brochures, menus, and one-page announcements, the layout is the message. Rebuilding it would throw away the design you (or your designer) made.
Preserve the design is an alternative pipeline for exactly those documents. Instead of re-authoring, it tags your PDF in place: the output renders pixel-for-pixel identical to what you uploaded, and underneath it carries a complete accessibility structure — headings, reading order, lists, image descriptions — meeting the PDF/UA-1 standard that screen readers rely on.
When you'll see the option
When you upload a PDF, we run a quick design analysis (a few seconds). If your document is a good candidate — a designed page with a real text layer and embedded fonts, or a scanned/flattened design where we can recognize the text from the image — you'll see a choice:
- Preserve the design — tag the PDF in place. You get back your file, visually unchanged, fully tagged.
- Convert to an accessible layout — the standard pipeline. Best for text-heavy documents, long reports, and anything where reflowing improves readability.
For visually rich pages (lots of graphics, short text) we lead with preserve-the-design; for lighter designs we recommend the accessible layout but still offer both. Documents that aren't candidates at all — long prose documents, plain scans without a designed layout — go straight to the standard pipeline with no extra step.
A few fillable form fields are fine — they get tagged too (with tooltips derived from their labels) and stay fillable, which no other pipeline preserves. If your document is mostly form fields, we'll also suggest Gevell Forms, which rebuilds it as an accessible form people can fill out online. If your document is digitally signed, you'll see a warning first: the tagged copy looks identical, but any modification invalidates the digital signature.
Pricing is the same either way: one page credit per page.
What you get back
- Your PDF, unchanged to the eye. We verify this mechanically: the output must render at least 99.95% pixel-identical to your upload, or we don't deliver it.
- Full tagging underneath: document title and language, heading structure, reading order, list markup, and AI-written image descriptions.
- Independent validation: every output is checked with veraPDF, the industry-standard PDF/UA validator, before it reaches you.
- A WCAG report documenting what was verified — useful for your own conformance records. See WCAG reports.
When the job finishes, the file shows Completed on your dashboard. Click it to open a summary of the verification results, any advisories, and download buttons for the tagged PDF and the report.
The summary lists every check we ran and its result — and is explicit about what we can't machine-verify: the reading order, the heading structure, and the wording of the AI-written image descriptions. Review each of those in the tagged PDF, then check it off on the file's page. Your attestation (who verified, and when) is recorded and becomes part of the WCAG report — useful when you need to document human review alongside the automated results.
Reviewing image descriptions
Every image in the tagged PDF carries an AI-written description. On the file's Summary tab, the Image descriptions section shows each image with a cropped preview and its current wording. Edit the text and select Save description to rewrite the description inside the tagged PDF — the visual design is never touched. Re-download the PDF afterwards to get the corrected file, then check off the image descriptions verification to record that a person reviewed the wording.
Advisories: things tagging can't fix
Preserving the design means we never change how your document looks — including things that are accessibility concerns but baked into the visuals:
- Small text. We flag text below 12 points. WCAG doesn't set a minimum font size, but small print is hard to read — especially on paper, where there's no zoom. It's common for the fine print to be the least readable part of a flyer.
- Low contrast. Text that doesn't meet the WCAG 2.1 AA contrast ratios (4.5:1 for normal text, 3:1 for large text) is flagged with the measured ratio.
These appear in your report as advisories so you can fix them in your design tool if you choose. They don't block delivery.
If verification doesn't fully pass
Occasionally a document tags cleanly but doesn't pass every PDF/UA rule — for example, a few text elements that couldn't be safely tagged. When that happens, nothing is delivered silently. You choose:
- Keep it anyway. You download the tagged PDF as-is, along with a report listing the specific gaps, and your page charge stands. Many teams take this option when the design matters more than the last few validation rules — the report gives you documentation of exactly what's outstanding.
- Convert to an accessible layout instead — no extra charge. Your original charge transfers, and the same document runs through the standard pipeline for free.
If the output fails our fidelity check instead — meaning it wouldn't look identical to your upload — we never deliver it, and your page credits are returned automatically. The same applies if a job fails outright or can't run: undelivered work is refunded without you needing to ask.
Good candidates
| Works great | Use the accessible layout instead |
|---|---|
| Event flyers and posters | Multi-page reports and papers |
| One-page handouts and announcements | Text-heavy newsletters |
| Menus, programs, postcards | Plain scans of prose documents |
| Brochures (up to 4 pages) | Form-heavy documents (try Gevell Forms) |
| Sign-up sheets with a few fillable fields | |
| Scanned or image-exported flyers (we recognize the text from the image) |
Exports from design tools like Canva, InDesign, and Illustrator are typically ideal. If the export flattened its text into an image — or the flyer is a scan — we can still preserve the design: the text is recognized from the page image and embedded as invisible reading text.
FAQ
Does it change my file's appearance at all? No — and we prove it on every job. The pipeline only wraps the existing content in accessibility structure; it never moves, restyles, or redraws anything. Output that isn't pixel-identical is not delivered.
My flyer's text is an image (scanned or flattened export). Can you still preserve the design? Usually, yes. When a flattened or scanned page still looks like a designed flyer, we recognize the text from the page image and embed it invisibly as reading text — the page itself stays untouched. Your report will note that the reading text came from image recognition, and if any line can't be recognized confidently we tell you instead of shipping a wrong transcription (you choose: keep it with the noted gaps, or convert to an accessible layout at no extra charge).
Why did my flyer not get the option? The most common reasons: it's longer than a few pages, or it's a plain scan without a designed layout (a scan that still looks like a designed flyer usually qualifies — we recognize the text from the image). Fillable form fields are not a blocker — a designed page with a handful of fields is a great candidate, and the fields stay fillable.
Is the tagged PDF really accessible? Yes — every delivered file passes veraPDF's PDF/UA-1 validation. PDF/UA is the machine-readable foundation; the accompanying report also notes WCAG-level advisories (like contrast and text size) that come from the visual design itself.