I think there are two things that are taken for granted by designers (the first of which I just observed yesterday in an email). First, you must define the instructional problem. I have seen designers jump in when management has stated there is a problem without a) confirming the problem exists or b) at least doing a goal analysis to obtain agreement on the outcomes which can also disrupt the plans. For example, I observed a case yesterday where the company was pushing very frequent training to their financial advisors and the advisors were resisting the training. The rationale I received was that the products are continually changing. I am not sure any type of analysis was done. It appears they were doing training because training was probably needed. In reality, a job aid or simply a bulletin might have been more effective and time and resource smart. Part of the problem may have been related the second issue.
Second, when I asked me students at the end of the design class which step of the process had little impact on their design and they might skip in future projects, it is almost always the learner analysis step. If you look at what we know about learner analysis and then how we treat the analysis in the strategy design it is weak. We have found that learning styles have no research foundation and do not have a role in the design of instruction. The aptitude-treatment interaction studies of the past century produced no useful heuristics. Basically, learner (and environmental analysis) tend to limit our designs. Thus, learner analysis has not worked out the way we thought it would, or at least the way my professors projected in the 1970’s.
Today, the learner analysis limits our design such as two hours of instruction starting one hour before work rather than 40 hours of instruction in one week. Or, not all students have access to that application or a laptop computer. Then, there are specific learner characteristics that can limit instruction such as eye sight, reading level, and prior knowledge (e.g., students with all levels of background knowledge). All these characteristics are important and must be accounted for when we design instruction. I have seen too many examples of inappropriate designs that failed to account for the learner. One of the classics stories of bad learner analysis was from a federal grant some 40 years ago. The design team created audio tapes for teachers of the deaf. Once they tried to implement the materials, they learned that a large number of the teachers were also deaf. Thus, learner analysis may not be as exciting as say a needs assessment or task analysis, we must still do a learner analysis even though it may limit what we can do, or to look at it in a positive way, it can create some great challenges for creative designs to address the instructional problem.