I would like to explain why PDF files on their own tend to reach a much smaller web audience.
One cannot assume that just publishing content on the Internet will result in your target audience finding it. In fact, without careful consideration of how that content is written, what form it is in, how it is marketed and what metadata it contains, it is unlikely that anyone will ever find it.
When it comes to PDF files and their find-ability via search engines, the video below, featuring Matt Cutts of Google, does a good job of explaining what Google needs to balance in order to serve up the most relevant search results. At the end of the video is a link to another video explaining some of the history behind why PDF search results need to be balanced with HTML content and why Google favours HTML content over PDF content.
Over the past four years of publishing content on unu.edu we have collected statistics that support our approach in which PDF documents are paired with shorter summary articles in HTML. These articles are written in a compelling style that is easier to understand by a wide audience, but links to the original full PDFs are also provided offering readers the chance to dig deeper into the topic. The result is an additional (and easily searchable) format to promote UNU’s PDF content.