What I would add to this post is this: for any system to gain any kind of traction you need to be aware of the barriers to engagement, especially if you previously had an underused LMS.
Factor in the fact that any system is there to enable and facilitate learning so really must have the user at the heart of it. To do this, you need to ensure that you do your homework.
- Research the potential users - what do they ewant, how would they use such a system and how did previous iterations fail. Segment user types to get as full a picture as you can - this should be ongoing
- Research the stakeholders - what are the business goals for this system, what will success look like and how will this be measured.
- Create user personas - imaginary (but based on reality) profiles of typical user types which look at overview of who that person is job role, tasks etc, look at user goals fro LMS and look at business goals - what that person is trying to achieve in their role. Also provide some general info - hobbies, online activity etc
- Design the system but include users in the process - take users from the research as they are likely to be more enagaged
- Design . . . test . . . build . . . test . . . and keep challenging your assumptions. And ensure you socialise your developments to your users so they get sight of developments and can provide a useful feedback loop.