Mike Lyons explains why optimizing individual parts fails: you need to optimize the whole system. Value stream mapping shows all steps to deliver customer value, revealing that waits between steps often dominate lead time. Eli Goldratt's Theory of Constraints says focus improvements on the bottleneck, then move to the next one.
Now, another important aspect of managing a system and thinking about that is looking at the entire system. I've seen a lot of projects where they have the entire project workflow and then they sort of sprinkle agile development. You know, like in the development phase, if you look at the whole project plan, it's sort of big upfront design. We're going to gather all the requirements, we'll spend months doing that. But then for dev and test, you guys be agile in there and do scrum or whatever, and then we're going to do big bang delivery on the back end of this thing. But what tends to happen is people start to ask the question, hey, why aren't we delivering faster? I mean, this whole agile thing is supposed to be twice the work in half the time. We have to think about the entire process for delivering value to our customers. The whole system. You can't just squeeze development and test and hope that it gets done faster. It's crucial to understand that improving individual parts of the system, what we might call local optimization, Think of a system like a chain. It's only as strong as its weakest link. In project management, we call this weak link the bottleneck or the constraint. I sometimes use those words interchangeably. But just as a roundabout improves traffic flow better than traditional traffic lights, focusing on your biggest bottleneck improves your entire process. More than fixing separate parts, this approach boosts the whole system's performance. Okay, but, so where do you start? A useful technique for seeing the whole system and understanding your entire process end to end is to create a value stream map. And I have an example for us to take a look at. Now, what we're looking at here is a very simplistic example of what you might go through. But a value stream map is a visual representation of all the steps that are required for you to deliver value to your customer. Each step of the process is represented as a box, and in that box you would describe the activity that's happening there. Now, once we have that in place, once we understand the steps that are required to deliver value to your customer, well, now we can start to add in some more information. We can start to use data to understand how much of our time is being spent at each of these steps in the process, as well as how much time is being spent in between the steps in the process. And inevitably, what I see over and over. And when we start to look at these steps, we find that most of our time is actually spent waiting the in between steps. When we look at this example here, if you're going to optimize this process, where would you focus your efforts? Well, as we look closely at the information, it looks like right here, we've got a four week waiting period between code and test and IV and V. Now IV and V, that's independent verification and validation. It's an unbiased third party assessment of the software development process and the resulting product. But, but that's a month that the work is sitting idle. Often our instincts though is to focus on the longer phase and try to reduce it. Like we look at code and test and just say, well, how can we get that to go faster? Let's hire more developers. How do we squeeze this four hours worth of work and get it down to two or three? But the reality is, even if we were able to cut this in half, the total amount of time it takes to get to the customer wouldn't really be impacted all that much. Instead, we need to focus on our longest wait time, that four week waiting period between code and test and IV and V. So when you think about optimization, it's important to look at the entire system and optimize for the whole. Now Eli Goldratt said any improvement anywhere except the bottleneck is an illusion. Focusing efforts on non constraint areas won't significantly improve the overall performance. Now here's some interesting news. Your gift for fixing the constraint. Well, it's the next constraint. Well, this is how you're going to maximize the flow of work through the system. You're going to map out your system, find the constraint, improve it and then find the next constraint. In fact, this is a recipe for managing the system. Now what I've been talking about here is rooted in the theory of constraints. Developed by Dr. Eliyahu M. Goldratt. He goes by Eli Goldratt. Well, these management concepts, they're powerful and they go well beyond what I've covered here. I encourage you to explore it further and apply these principles in your work. It could be a game changer in how you approach process improvement and overall system optimization.
