Auto-adjustment of finish date

anthony0988,

First, let me answer your two questions directly.

1. When you schedule from the finish date Project sets a constraint of "as late as possible". Thus when your first task finished early, the second task was constrained to start as late as possible such that it met the 5/24/24 project finish date.

2. Is Project not smart enough to know your plan could finish earlier? No, Project is "dumb" it doesn't "know" anything. It is just a tool that responds to user inputs and creates schedules based on it's scheduling algorithms. It's up to the user to understand how Project works and then it can be used very effectively to help create and manage schedules.

Tempting as it may seem I suggest you NOT set your project to schedule from the finish. Being new to Project is challenging enough but reverse scheduling will make things all the more difficult, as perhaps you are finding out.

Nearly everybody who creates a schedule has a "do or die" finish date, nothing unusual there. I'd start by first resetting your project to schedule from the project start date even though at this point you may not know when that needs to be. Often that start is driven by a contract start or some other event that "starts the ball rolling". But even if all you have is a finish date I'd start by laying out all the tasks necessary to accomplish the end goal with their estimated durations. Keep in mind that in Project duration is the time span during which each task is performed. Duration is NOT the same as work. Work is the amount of effort from one or more resources to accomplish a task. If a single resource is assigned full time to each task, then work and duration will be equal.

Make sure all tasks as auto-scheduled. If you want to do some rough up front planning, you can set up your plan as manually scheduled. That will allow you enter dates and quasi dates (i.e. "next week") into the Start and Finish fields. But always use auto-scheduling to create the final plan and throughout execution.

With all your tasks and their estimated durations entered, link those tasks in a normal sequence of performance, usually with a finish-to-start-dependency. Do NOT manually enter start or finish dates as that sets a constraint and thwarts Project's scheduling engine from dynamically scheduling the plan.

Once you have all the tasks defined, assign the resources necessary to preform those tasks. Work type resources may be human labor or machine labor. Normally, resources are assigned at 100% but there are many factors that determine how much and how many resources may be necessary to perform a given task.

The project calendar is an essential element. By default the Standard calendar (5-8) with a start time of 8:00 AM is the project calendar but if your company works something other an a 8 hour day and/or has a different start time (e.g. 7:00 AM) then you need to create a custom calendar. Do NOT directly modify the Standard calendar, rather, make a custom calendar, perhaps based on the Standard if appropriate, for your project. Calendars have a major impact on how your plan lays out, so set up the calendar before defining your tasks.

Is you head spinning yet? If not, it should be. Project is not an intuitive application and the learning curve can be steep, particularly if you've never working in a project management environment before.

This treatise is perhaps a little jumbled but once you get your plan laid out, does it meet the required end date? If not, then make adjustments (i.e. start earlier, re-arrange the sequence of tasks if possible, add more resources, etc.). Iterate the plan until it "works". Before executing the plan be sure to set a baseline that that form the basis for measuring performance.

Is your brain full yet?

John