A Schedule is Not a Still Life
Schedules need to move. Two key tests for whether your schedule will work for you or against you.
Scheduling is an art, but it’s also a science. A good Critical Path Method schedule is a network of interrelated tasks that captures duration, constraints, and relationships, and spreads them all across available work time and resources. It is also a huge pile of assumptions, a central battleground in disputes, and a fragile document made using specialized software that needs constant feeding and care to be relevant. And when you get down to it, everyone just wants to know (1) how it’s going, and (2) when things will be done.
CPM schedules evolved into their current form between the 1940s and 1960s, with contributions from the US Navy, the Manhattan Project, DuPont, Booz Allen Hamilton, and Remington Rand. They were pretty revolutionary when they were first introduced, and nobody’s gotten much traction with any other way of representing a complex schedule since. Most importantly, the methodology is enshrined in various standard contracts across the world, which require CPM schedules and Gantt charts. Even if a contractor wanted to use a different method for the work itself — Lean Construction techniques are popular for a reason — a CPM schedule still usually has to be used for reporting, and it will almost certainly form the basis for any dispute over delays.
Is Your Schedule a Snapshot?
This is why it is so important to have a schedule that is living. Too many schedules are built with one goal in mind: a PDF Gantt chart where the bars line up where they’re supposed to. The task durations are tweaked so that the major milestones come out correctly, and the relationships are just there so that the right activities show up as critical on the chart.
This kind of schedule is a still life. It’s a snapshot of the current plan and fundamentally a reflection of the art of the scheduler. The scheduler is the only one who can tell if the major milestones are “correct” and if the “right” activities are critical, because the timing and criticality are not really being generated from the bottom-up relationships in the schedule.
That also means that if anything changes, adjusting tasks in the schedule will not translate into the real effect on other tasks or even on the end date.
This isn’t helpful for anyone. At the beginning of a project, the baseline schedule establishes a shared reality: the key milestones are there, the critical path tells everyone what tasks are the most important to get done on time, and progress on tasks means progress on the project. If the baseline schedule isn’t solid, that shared reality is built on sand. Any changes will cause it to fall right over. It will be harder to understand the impact of delays, harder to keep the schedule reflecting reality, and harder to keep trust in the relationship between owner and contractor.
The Only Constant in Projects is Change
That’s a problem, because pretty much every project is going to stray from the schedule at some point. Risks materialize, materials are late, crews under- or over-achieve, tasks have to be resequenced, things happen. With a living schedule as a starting point, it’s possible to make these changes and have the schedule recalculations be meaningful. You should be able to learn things from your schedule.
With a “still life” schedule, it’s much more work, because the whole schedule might have to be changed to reflect the current plan. In addition, those changes might be very difficult to explain. Why would delays to excavation require changes to the relationship between millwork and painting? Well, because the original relationship was set up solely to put those tasks on the critical path, and they’re intuitively still on the critical path, but the calculations don’t show it anymore. The new critical path has to be forced into place — and the owner now doesn’t trust it.
Test Your Schedules
So how can you tell if your schedule, or your contractor’s schedule, is a still life? There are lots of tools out there for doing analysis of schedule quality, some automated and some not. Primavera Risk Analysis’s Schedule Check tool, Deltek Acumen Fuse®, and the US Defense Contract Management Agency’s 14-point analysis are a few. These tools apply various tests to attempt to identify whether a schedule is probably a good reflection of the real relationships, or whether it’s been built to produce a nice-looking Gantt chart. If you don’t have these tools, or don’t want to go through the full manual process, here are two of the most useful tests:
- Do all of the tasks have real relationships?
- Will delay to tasks on the critical path show up in the end date?
Connectedness
The tasks in your schedule should be connected to other tasks in your schedule; the other name for a CPM schedule is a schedule network. It’s like a fishing net. If you tug on one string, it should have effects on the rest of the net. Tasks which aren’t fully connected to other tasks, because they’re scheduled to start or finish on a certain date, or they’re only connected at the beginning or end, are like cutting holes in the net or nailing the net to a wall. They prevent effects from propagating the way they should. If there are lots of tasks in the schedule that aren’t connected through schedule logic, the schedule won’t behave properly and you won’t be able to really even apply the other tests. You don’t have a fishing net, you have a collection of scraps. The connections are the place to focus first.
While you’re looking at the connections, it’s also a good idea to think about ways to turn your relationships into “finish to start” rather than finish to finish or start to start — finish to start relationships are the most predictable and robust in a network, while finish to finish or start to start relationships can have unintended side effects once changes begin to be introduced. High amounts of lag and lead can also be unpredictable.
Criticality
Once the tasks are connected and you have a critical path, you can apply the second test. The entire purpose of a critical path is to indicate the longest path through a schedule. Any delay to tasks on the critical path should theoretically result in delay to the end date of the project. The question is… does it? It’s quite possible to build a schedule where delay to tasks on the critical path doesn’t delay the end date. One of the most common reasons is too many fixed-date constraints, but excessive lag and lead or overuse of SS and FF relationships can also cause problems. This is a simple but extremely effective way to find out whether your schedule will function the way it should if there are changes: change it and see! The end date should move by the same amount added to each task. If it doesn’t, find out why.
Bringing your Schedule to Life
The two previous tests should have identified whether the schedule is a still life or a living schedule, and they should also have helped identify what to do to fix it. A truly living schedule will be fully connected and responsive to changes on the critical path. It will have almost all Finish to Start relationships, very few fixed dates, little lag, and little lead.
Getting to that state may require breaking tasks into pieces, introducing new tasks, and thinking hard about the real relationships between tasks. It can be a fair amount of work to turn a still life into a living tool, but the payoff is the ability to report realistic progress, understand and demonstrate the effects of delay, explain changes reasonably, and restore everyone’s confidence. That’s worth putting in the time.
