On the topic of inferring the value of a document from the number of
downloads, I would like to share this true story.

When I started my first online community, devoted to meeting the
needs of educators at HP, one of the more popular topics was reviews
of training videos and CBTs. We even licensed and uploaded reviews
from Bill Ellet's newsletter "Training Media Review."

One article we got from TMR was a review of a training video on
sexual harrassment. The article was titled "Sex at Work," so that's
the title we used when copying it to the online community.

The number of page views of that particular posting FAR surpassed
any other contribution to the community.

This led me to understand that "reads" and "downloads" in many
contexts are driven almost entirely by the item's TITLE, not the
quality of the content.

If you are looking for a very simple metric that correlates with the
quality of the CONTENT, I suggest that you might want to think about
counting the number of times the document is revised. When a
posting goes from version 1.0 to 2.0 to 3.0 you know that it's still
alive and that someone cares about improving it. (This can, of
course, be gamed, but I think the point is still valid.)

Another sign of aliveness is the number of replies to the base
posting. This metric can be used to gauge the aliveness of the
community as a whole, in addition to the aliveness of a particular
posting.

Cheers,
Bruce Karney
http://km-experts.com