Mary Abraham writes about how metrics can drive the wrong behavior (or the wrong conclusions) in The Metrics Mess.
[T]ake a moment to consider whether your efforts to measure the wrong thing are leading you into unproductive activity. Don't focus on bulk -- focus on impact. For example, counting how many times a particular document is opened via your portal or document management system may be interesting but not helpful. What you really want to know is how many times was it opened and actually used? And, how often was it exactly the thing the user was searching for? In the latter two cases, you learn much more about the quality of your content and the quality of your search engine.
Of course, the place many organizations struggle is with that measure of "impact." So they develop proxies for it, such as number of times accessed.
On the wider web, the proxy for "impact" is the number of links to a given item. If lots of people link to it, that is an interpretation of "impact" or importance. (Google calls this pagerank, of course.) But how do you measure the impact of the FMLA policy document, or a sample document, or ... With most corporate systems, there is no sensible way to measure impact in this way. Beyond "are people seeing this content" in general, I suspect there isn't much point in trying to measure things this way.
But, how does an organization get beyond this? I suggest the idea of reinforcing the use of this stuff in the business. Ask your colleagues if they've shared their recent successes and failures. Make it part of the regular process around reviewing projects: Ask people who they asked and what resources they consulted before initiating a project. If you are feeling autocratic, turn off attachments for internal email -- being sure there are mechanisms for publishing "stuff" for the internal population.