Google indexes many non-HTML formats: PDF, Word, Excel, PowerPoint, plain text, RTF and several others. They appear in results with a file-type label.
Practical implications
- PDFs cannot carry a
<meta>robots tag, use theX-Robots-TagHTTP header instead. - Canonical for a PDF is set via the
Link:HTTP header. - A PDF title comes from its document properties, not the filename. Set it.
- Text in a scanned image PDF is not readable unless the file has an OCR text layer.
Alien Road recommends: publish an HTML version of anything important and keep the PDF as a download. PDFs rank, but they are poor on mobile, hard to update, and carry none of your navigation or calls to action. We treat “PDF-only” content as a conversion problem before an SEO one.