Paper Trail: Auditing the PDFs Doing a Web Page's Job
There's a familiar sort of UK business website where the most important content isn't really on the website at all. The service descriptions live in a nine-page brochure exported from InDesign in 2019. The price list is a PDF behind a link labelled 'Downloads'. The menu, if you're a venue, is a scan of the laminated version, photographed slightly at an angle.
The site itself is a shell: logo, hero image, phone number, and a button that says 'Download our brochure'. It feels finished because someone made it, but what it actually does is ask a stranger to complete a small admin task before they're allowed to decide whether to call you.
That's the trap. A PDF is a download, not a destination. And when it becomes your front door, you pay for it in four different currencies at once.
A download is not a landing page
When someone taps a PDF link on a phone — which is how most of your traffic arrives — they leave your site. The file opens in a viewer with its own toolbar, its own zoom, and its own back button that dumps them wherever they came from rather than back to your page. To read a print-layout document on a 6-inch screen they'll be pinching, dragging and squinting, hunting for a phone number that's typeset in 8pt on page 11.
A web page, by contrast, reflows. The headline is readable without zoom. The call button is thumb-width. The next step is visible. These sound like small differences. They're the difference between a readable proposition and a homework assignment.
Four bills you're quietly paying
Visibility. Google can read PDFs, but they're second-class citizens. You don't control the title tag or the meta description in any meaningful way, there's no internal linking, no schema, no clean heading structure, and the snippet Google shows is often lifted from a contents page or a stock legal paragraph on page two. In an era where answer engines want tidy, quotable chunks of text, a fixed-layout PDF gives them almost nothing to work with. Your competitors' HTML service pages will outrank it every time.
Accessibility. WCAG 2.2 AA asks that content reflow at 320px wide without two-direction scrolling. A print-designed PDF simply can't do that. Most aren't properly tagged either, so a screen reader announces them as an undifferentiated wall of text. If you work with public sector clients, remember that their accessibility obligations tend to follow the documents they publish or link to — a point worth raising before someone else raises it for you.
Measurement. Unless somebody has deliberately set up an event, opening a PDF doesn't register as a page view. Your analytics look quiet, so you conclude the page is unpopular or the channel doesn't work — when in reality the content is being read in a viewer you can't see. You're making decisions on a partial picture.
Conversion. A PDF can't host a form. It can't offer click-to-call. It can't personalise, A/B test, or show a live price. And every time something changes, someone has to re-export the file, re-upload it, and hope the right version went live. Version control on a shared drive is not a content strategy.
The audit: one spreadsheet, one afternoon
You don't need a consultant for this. You need a list and a critical eye.
- Find every PDF. Run a crawl with a tool like Screaming Frog, or search
site:yourdomain.co.uk filetype:pdfin Google. Export the lot. - Log where each one is linked from. Main navigation, homepage, service pages, footer, a QR code on the van. Anything in the nav or above the fold is urgent.
- Test each one on a phone. Actually do it, on mobile data, not office Wi-Fi. Time how long it takes to find a phone number or a price.
- Check the tracking. Is a
file_downloadevent firing in GA4? Enhanced measurement usually catches it, but confirm — and add values for the documents that matter commercially. - Ask one question of each file. Does this PDF contain the only explanation of a service? If yes, flag it for conversion to HTML. If it's a supporting document, leave it alone.
When a PDF is genuinely the right tool
None of this means binning your PDFs. They're excellent for things people want to keep, print or forward: a takeaway menu with prices, a technical spec sheet, a timetable, a compliance certificate, a tender pack, a conference programme. Anything where page fidelity matters, or where a customer has explicitly asked for a file.
The rule of thumb is simple: the PDF should be the takeaway, not the restaurant. And when you do offer one, label it honestly — 'Download the 2026 price list (PDF, 1.8MB)' tells people what they're getting and stops the surprise.
Turning the brochure into a page
Pick the three or four sections of the PDF that answer the questions people actually ask before they buy. Write them up as proper headings and short paragraphs, one idea per section. Put prices and specifications in a real HTML table rather than an image of a table, so it reflows and can be read by search engines. Add one clear next step above the fold — call, form, or booking link — and keep the download as a secondary option below it.
Then do the boring housekeeping: redirect the old PDF URL to the new page so printed QR codes and old links still land somewhere useful, keep the file itself available at a fresh URL, and track downloads as an event so you can see whether people still want the full document.
The short version
Your website's job is to answer a question quickly on a small screen. A PDF's job is to be a document. Most SMEs need both, in that order. If the only route into your services is a button marked 'Download our brochure', you're asking a stranger to do paperwork before they've decided whether to ring you — and plenty of them will simply go back to the search results instead.