The heart of the problem is the core of the plan

The most critical part to any plan is the objective. Every plan has to have an clear, consise and explicit objective, or the end is never reached. This objective needs to be measurable. So, "increase revenues in Q1" is not an appropriate objective. It isn't very difficult to make it measurable, though. "Increase revenues 10% in Q1" is an appropriate objective. Not all goals are that simple, though. In the creation of a new piece of software, or migration of multiple datacenters into one, the objective needs to be more fully defined, so that should be your first step.

The definition of the objective is often called a requirements document and should be the first step in planning. It should include the following elements:

1. Background: What is the current situation / landscape
2. What is the overarching goal or objective
3. Explicit sub-requirements for reaching that objective
4. A picture of the future - what will be different from current situation / landscape

Many methodologies have intricate templates for publishing these documents. As long as they contain and capture the above, choose your flavor.

The creation of this document should include a critical analysis of the objective, the sub-requirements and the outcome against the intended goal. If these are all consistent, then the next test is to make sure we are still on the right track. WHAT WILL IT COST? This is a critical component and not one to take lightly. The measurement of the cost is always a bit fuzzy, but it should be estimated up front and measured against gain from the intended goal. If we expect positive return on investment (we are going to derive "value" from the product now or in the future), then we proceed with the plan. If not, now is the time to kill it!

The next step is to break down the big pieces into smaller tasks (work breakdown structure) and assign out those tasks. There are multiple tools one can use to record, track and measure the WBS (work breakdown structure). Again, whatever your choice of methodology, just keep the big picture in mind. It all comes down to holding people accountable for doing what they said they would do. I have known project managers who are supremely effective at running complex projects within an excel spreadsheet; and others who couldn't manage a project with MEGA tools provided by MEGA software companies.

Once you have a plan, it's all about communication. Everyone should understand the big goal, their part in it, and why they will benefit from it. It seems like that wouldn't be a huge deal, but you would be surprised at how many times a project will stall because its resources lose sight of "the ball".

MANAGE RISK: The biggest risk to the objective is not managing and communicating risk. I would consider this the project managers primary role. Creating a work breakdown structure and checking the boxes as things are finished is actually pretty simple. A GREAT project manager will constantly be acquiring, absorbing and analyzing information from resources, external issues, budget, etc. and build a risk profile and communicate that to stakeholders.

Well, I've oversimplified I'm sure. As with anything else, there are thousands of intricacies in the execution, but I hope I've hit some of the more critical elements involved in successful planning. People have written entire libraries on the subject, so I can't do it justice here...except to offer what's important from my perspective.