Problems arise in this area for two main reasons . People tend to underestimate the time it takes to complete tasks regardless of past experience so estimates of time tend to be optimistic at best. Often schedules are set by people who are not actively involved in project work and therefore have unrealistic expectations of completion. Whether due to overly optimistic time estimates or unrealistic expectations the target times will always be exceeded.

Therefore time estimates for tasks must always be treated as just that estimates. When project plans are based on inaccurate estimates the first task that takes longer than expected throws the entire plan out of whack or requires bringing new people on to the project to make up for lost time. In multi-project management a delay in one project can quickly affect the schedule and/or resources of the others. This is how you manage the schedules of several Specific Database projects Plan buffer times and larger time windows in the project plans which absorb possible miscalculations. All you have to do is increase the estimates for tasks by a certain factor for example by planning five days for all tasks that are likely to take three days.

If you know that ad hoc requests take up of your teams time you can increase their time on tasks by . Such a buffer makes it possible to stick to the project plan despite the almost inevitable delays and changes in the process - and above all to prevent delays in one project from affecting the others. For example if a QA team has project A scheduled one week and project B the following week but project A is delayed by a few days the following possibilities exist Project As review is postponed until Project B has been reviewed. This means that A will only be completed two weeks later than planned. Both project A and project B are pushed back by one week and are therefore only finished one week later.