Optimising PDF Files for Web Delivery

← Back to Blog

Why Web-Optimised PDFs Matter

PDF files created for print production are frequently much larger than necessary for web delivery. A press-quality PDF may embed images at 300 dpi or higher, include full font sets, retain print-specific metadata, and store embedded thumbnails and colour profiles relevant only to print workflows. Delivering such files over the web imposes unnecessary load time on users, wastes bandwidth for both server and client, and degrades the experience particularly on mobile connections. A well-optimised web PDF typically achieves file size reductions of 60–80% from a print original with no perceptible loss of quality at screen resolutions.

Linearisation and Fast Web View

A standard PDF file is structured so that the cross-reference table — which maps object numbers to byte offsets — is written at the end of the file. A browser or viewer must download the entire file before it can locate and render the first page. Linearisation (known in Acrobat as Fast Web View) restructures the file so that the objects needed to display the first page appear at the beginning of the byte stream. A web server that supports byte-range requests can then send the first page’s data before the rest of the file has finished downloading, allowing the document to begin rendering immediately.

Linearisation is especially valuable for long documents where users often need only the first few pages. In Acrobat Pro, enable it via File > Save As Other > Optimized PDF and ensure the “Optimise for Fast Web View” option is checked. Verify the result in File > Properties, where Fast Web View should display “Yes”.

Image Resolution Targets for Web vs Print

Images are the dominant contributor to PDF file size in most documents. Screens cannot display more than approximately 96 dpi of effective resolution at normal viewing distances, so embedding images at 300 dpi provides roughly nine times more data than the screen can render — all of it wasted. The recommended targets for web-delivered PDFs are:

• Colour and greyscale photographs: 72–96 dpi for screen-only use; 150 dpi if end-user printing at home is expected.
• Monochrome (bilevel) images: 300 dpi is acceptable for scanned text even in web PDFs, as JBIG2 and CCITT compression keep file sizes modest at this resolution.
• Avoid downsampling below 72 dpi — results appear blurry at normal zoom levels and undermine the document’s professional appearance.

JPEG vs ZIP Compression for Images

The choice of compression algorithm for embedded images significantly affects both file size and visual quality:

• JPEG (lossy): Best for photographs and continuous-tone images. JPEG discards high-frequency detail that is less perceptible to the eye. A medium quality setting (typically 60–75 on a 0–100 scale) produces visually excellent results at a fraction of the uncompressed size. Avoid JPEG for images containing sharp edges, text, or flat areas of colour, where compression artefacts are clearly visible.
• ZIP / Deflate (lossless): Best for screenshots, diagrams, logos, and any image with sharp colour boundaries or text. No data is discarded, so quality is perfectly preserved. File sizes are larger than JPEG for photographs but appropriate for non-photographic content.

Acrobat’s PDF Optimizer allows different compression settings to be applied automatically based on image type, which is the recommended approach for mixed-content documents.

Font Subsetting

Embedded fonts can add substantially to PDF file size, especially when full font families are embedded. Font subsetting embeds only the glyphs (individual characters) actually used in the document, rather than the complete font file. A subset of a font used for body text might reduce the font’s contribution to file size by 80–90%. In Acrobat, set the subsetting threshold to 100% to ensure subsetting is always applied regardless of how many characters are used. For documents using decorative typefaces for only a few headline characters, consider converting text to outlines and removing the embedded font data entirely.

Removing Unused Resources

PDFs created through complex authoring workflows accumulate data that serves no purpose for web readers:

• Embedded thumbnails: Acrobat and InDesign can embed pre-rendered page thumbnails for each page. Modern PDF viewers generate thumbnails on demand and do not need them pre-embedded — remove them to save space.
• Duplicate content streams: Some authoring tools create redundant copies of image or font data. Acrobat’s Clean Up panel in PDF Optimizer removes these duplicates automatically.
• Unused metadata: Custom XMP metadata namespaces added by authoring tools, version history records, and proprietary embedded data can be stripped safely. Keep the standard Title and Description fields for accessibility and SEO.
• Review comments and annotations: Any review markup added during production should be removed before web publication.
• Embedded file attachments: Remove attached files unless they are integral to the document’s purpose.
• Optional content layers: Flatten layers if only a single variant of the document is published.

Using the PDF Optimizer Tool

Acrobat Pro’s PDF Optimizer (File > Save As Other > Optimized PDF) provides granular control over all the techniques described above. Before adjusting settings, click Audit space usage to see a breakdown of what is consuming space in the specific file — this allows effort to be focused where it will have the most impact. Work through the panels in sequence: Images, Fonts, Transparency, Discard Objects, Discard User Data, Clean Up. Save the configuration as a named preset to ensure consistent results across future documents.

For a one-step option with less control, File > Save As Other > Reduced Size PDF applies a standard optimisation profile suitable for many routine cases.