Waterfall is a classical sequential (non-iterative) approach used in large-scale packaged solution implementations globally. This has been used for initiatives/programs that are big, span multiple businesses, have a large number of requirements, involve complex single-site deployments, developed by global/multi-site organizations, and so on.

This approach is simple to understand and has checkpoints/phases that must be completed before the next phase can start.

The following image shows the Microsoft Sure Step methodology with discrete phases and their deliverables in order to give an end-to-end glimpse of the waterfall methodology:

Please check if Sure Step is available or has been stopped by Microsoft, as there are a lot of changes happening in the cloud world.

As shown in the preceding image, there are predefined stages and their associated deliverables. In a typical waterfall implementation, the following phases are expected:

• Diagnostic: This is primarily the presales phase, or the requirement gathering phase.
• Analysis: This is the phase where an advisor/partner understands the requirements and maps them with the solution. During this phase, the extent of solution fitment will be known and the decision to customize, buy, or build the solution verticals are decided.
• Design: This is the phase where a solution blueprint needs to be prepared and which in turn needs to be supported with other designs, such as functional, technical, data migration, security, and so on.
• Development: In this phase, all the gaps for which design documents are prepared in the earlier phase are undertaken for development and unit testing. Some implementations perform functional testing as part of this phase or in the subsequent phase.
• Deployment: This is the solution validation phase, where the solution is to be tested to fit the business requirements of the initiative. Preparation for go-live also happens in this phase, and so do the activities and deliverables related to data migration, security, training, and so on. 
• Operation: This is the phase after go-live, often known as support.

Let's now explore some key attributes of this methodology:

• This methodology is highly efficient when the requirements are very clear, fixed, and well-articulated. The scope for ambiguity must be less, as it would reduce the chances of success.
• Sometimes, implementations allow phases to run in parallel.
• There are some shortcomings of this methodology and they must be checked all the time:
  • Ensure that the concept stage is finalized and signed off, as once the system goes into the testing phase, it is very difficult to go back and change anything
  • The entire solution remains a black box and is seen only towards the later phase in the methodology lifecycle.
• Reconsider your methodology when the requirements are to be elaborated out or when the program is an on-going initiative.
  • There can be huge risks, assumptions, and uncertainties involved due to the inherent nature of the methodology.