Ithuta JV

Source: TenStep Article by Tom Mochal

Summary
You are right to be concerned about the unknowns associated with long projects. However, there are a number of techniques that can be used to make you feel more comfortable. Being comfortable does not imply that you know everything. Being comfortable means that you have taken your best shot at laying the project out as best you can, and then relying on a good set of communication processes, risk management and schedule management to discover the detailed nature of the project as you progress.

...

Many times you are asked to manage a long project that starts off with a lot of uncertainty. You may not be 100% sure of the resource requirements, the final deliverables, the cost, schedule, etc. The team members may not be allocated full-time to your project. They may be either assigned to other projects, or else to the support of production applications. There may be many risks. Part of the planning process involves deciding how to manage the unknowns.

The good news is that project managers deal with the uncertainty associated with large projects all the time, and it is very likely that you can be successful. You can imagine projects with dozens (or hundreds) of workers, and very long timeframes. If those project managers can be successful, you can too.
Okay, so what do you do? Here are a couple suggestions for you to consider, depending on what you think will work best in your situation.

Break the work into smaller pieces
The first thing to do with a long project is to break it down into smaller pieces if possible. For example, let’s say you have a traditional waterfall type of project. Although you are unclear about the work to be done in the next year, you should at least know what you will need to do over the next few months. You are probably going to start in a requirements gathering process. Instead of defining a one year project, start by defining a project that will cover only the analysis phase. After that project, you can redefine and estimate the remainder of the work. If you still feel uncomfortable doing that, then perhaps you can create a project that just covers the design phase. Ultimately, you may complete the work in three or four smaller projects instead of one large one, but you will get there nonetheless. You will also be able to more easily plan and manage each of the shorter projects.

Provide less detail as the planning horizon gets further out
Many organizations are not structured in a way that allows you to break a large project into a set of smaller ones. Many companies only want to pay for one project, and track one project. If you break it up into pieces, people get confused about what you are doing.
The next idea is to estimate and plan the work for the entire timeframe, but just understand that there will be less detail the further out in the future you get. Again, you should have a firm and detailed workplan for the next three months, but then the planning will be at a higher and higher level. You have a framework for completing the project, but only the short-term activities are planned out in detail. This is probably the approach most project managers take on long projects.
Of course, you cannot leave everything at the high level. Every month you need to re-plan the project, validating the detailed work for the next two months, and then building the details for the third month out. This makes sure that you always have a three-month detailed planning window, as well as filling in more and more detail for the outer months. If the detailed planning leads you to believe that you will not hit your deadline or your budget, attempt to resolve the workplan situation immediately or raise this possibility as a potential risk.
If you are concerned about the availability of resources, this approach should work. When you create the initial workplan, you can start to communicate regarding the specific people you will need in the next three months, and the general types of people you will need further out. If you keep a detailed three-month planning window, you will be able to give the resources managers up to three months lead-time once you finally nail down exactly whom you need. If resources are not available, you also have up to three months to escalate the problem or to look for alternatives. Two to three months notice should be enough time to manage through any of these resource scenarios

Use multiple estimating techniques
The classic estimating technique is to build a work breakdown structure, estimate the work associates with the lowest level activities and then add everything back up for the final overall estimate. This approach does not work well when you are not sure exactly what the work is a long way into the future. Fortunately, there are other estimating techniques that will help you cross-check your estimated effort, cost and duration. Fist, you can rely on outside experts to review your Project Definition and workplan to see if they think your estimates are reasonable. Second, you can see whether there have been similar projects in your company where you can review the prior workplan and estimates to see how they line up with your project. Third, you can use industry guidelines to create overall estimates based on how much time you think the analysis phase will take. If you find estimating guidelines, for instance, that say that the analysis phase of a project with your characteristic takes 28% of the entire project, then you can provide a high level estimate of the entire project based on your detailed estimate for the analysis phase of your project.

Risk management
Generally, the problem with long projects is that there are many things that can happen in the future that you do not know about today. In other words, there are potential risks in the future. We have already discussed some ways to deal with the uncertainty of schedule and effort estimating risks. There are other risks as well. In your case, for instance, there is a risk that resources you need in the future will not be available when you need them.

All of these risks can be identified, and then a specific plan can be put into place to mitigate the risk and ensure it does not happen. Every month, you would update the risk plan to ensure known risks are being managed and new risks are identified. If you think there is risk associated with any aspect of the project in the future, identify it and mitigate it.