The solution's structure reflects several key trade-offs regarding content and how it will be distributed to users:

• All-inclusive vs. selective.
  If the volume of information grows too large, the system will become slow and unwieldy, and people won't use it. One company widely known for its early success with knowledge management ran into trouble when users became overwhelmed with too many databases, too many documents, too many categories -- and couldn't find the information they needed. Today, the company no longer dumps entire contents of articles and reports into the system. It limits on-line content to abstracts, and directs interested readers to the authors for more information.
• Just-in-time vs. just-in-case.
  Knowledge is more valuable when it is delivered at the moment it's needed -- "just in time" -- rather than being available at all times -- "just in case." Just-in-case systems devalue knowledge, as people become accustomed to receiving information that's not relevant to their jobs. They may ignore the messages, or spend valuable time following threads that are interesting but not germane to their work.
• Push vs. pull.
  Under a "pull" system, users proactively seek knowledge when they need it; a "push" system distributes knowledge to its audience, often pushing it through customized filters. Pull systems don't distract users with unwanted updates, but require user initiative. Push systems get information onto users' desktops, and thus may be more likely to influence decision-makers, but they work best when they correctly anticipate users' needs, and don't push unnecessary knowledge on people.