Spot Delivery Problems Early Part 2: Fundamentals
How does a project get to be a year behind schedule? One day at a time.
In part 1, I talked about governance. If the owners of the project are not governing the basic behaviors to execute well, your risk of delivery is rising. If you haven’t read it, go back before your read this post.
Just a short reminder: I’ve written before about the most important measurements in projects and the behavior you need to deliver on time: proactive and speedy resolution of problems. These metrics are based on the premise that behavior is the precursor to results. If you want to know if you’re going to get the results, you should measure the behaviors.
In this post, let’s talk about the behavior to deliver on time; or why projects are late.
Someone, perhaps it was Elon Musk, says your car is the most under-utilized resource you have. Only 15% of the time. It’s typically the most expensive asset people have, after a house. It’s sitting. Not moving. In your garage. In your parking lot at work. It’s a pretty expensive convenience.
Like your car, projects are sitting most of the time, too. At least half the time. You could be going faster, but your project is waiting. Considering the value of a project, that waiting is quite expensive. I worked on an oilfield development project where each day early or late meant over $4mm in revenue per day.
Maybe you don’t have projects that are that have that sort of impact, but the point is that the earlier a project is completed, the sooner you receive the benefits of that project. Most of the time the project is waiting, why wait?
You can know – early – if your project is waiting around. Projects that are going to be late are comprised of tasks that are slow to complete. If you measure the rate of task completion, you can get sense of how fast your project is moving. You can measure the duration of tasks, too. If the task duration is increasing, your project is slowing down.
If you are doing projects over and over, you could measure the rate of completions. That would give you a sense of how your process is delivering, but it’s always after the fact. It’s not predictive. You must look at the actions that are needed to increase the rate.
The rate of task completions will give you a sense of how all work is proceeding, so you look at the rate and the accumulation of task completions. If tasks are completing quickly, that means that in general, waiting is minimized and we’ve achieved good flow.
The two elements give you a different look at completion velocity. One gives you an absolute number, the other gives you a sense of acceleration. If the slope of task completions over time is 45 degrees, you’re holding steady. Less than 45 degrees, you’re slowing. Greater than 45 degrees, you’re speeding up.
You can see that during part of the project things slowed down. A lot! Then things sped up again. Why is that? This will have a big effect on your delivery. The slowdown tells you that your project delivery risk is rising! Managing delivery risk is the PM’s job, isn’t it?
That doesn’t tell the whole story. We know the rate of the system is determined by the rate at the constraint. This graph tells us about all the project tasks. It could be, that the constraint isn’t working very fast, and all the non-constraints are working ahead. When that happens, your project delivery risk could still be risking and you wouldn’t know. So you must look at the throughput (completions) for the system versus the constraint.
In the example above, the constraint (bottleneck1 Material Processing didn’t complete anything for four weeks, then started completing more and more. The red line indicates the target throughput for this resource, 2 per week. On the right, the average is a bit less than target. So that means that this portfolio team is completing slightly fewer projects per week than planned. Schedule delivery risk is rising.
What about completions, once the projects pass through the constraint, do they get completed?
A similar wave effect as the constraint operation. Very little at the start, then a lot. We can safely say that that output at the constraint resource is a predictor of the output of the overall system. The operations geek in me is also wondering about the hockey stick effect. Hmm. Looks like something is broken there.
Knowing, measuring, and managing the precursor behaviors as measured by task completion velocity is critical to managing the schedule risk. Our teams look at overall completion velocity to see we’re going faster or slower and we look at the constraint completions as the predictor of system output.
Few (if any) project management systems have this sort of intelligence built into them, but VISUM visual project management software does. That’s why we built it. Other tools give you data, but none give you information you can act on. VISUM is built around the ViewPoint visual project management (execution) methodology that takes the best practices from PMBOK, Agile, Lean, and the Theory of Constraints to give you a consistent, scalable method for executing projects on time.
Check out how one team used these principles to rescue their project and cut project lead time by 27% in this Knowledge Work Case Study
Check out the ViewPoint Methodology here: http://pinnacle-strategies.com/services/project-management-consulting/viewpoint-visual-project-management/