Here's a common scenario. We need to make a relatively simple change to an application less than an hour of development time. We make the change but weeks later it hasn't been deployed to production. Customers start to get frustrated. An embarrassed manager escalates the issue, going to team after team and insisting that they prioritize the work that must be done. Finally, the change is pushed through and made available to our customers. What's going on here? In a word, handoffs. When we see long lead times, one of the underlying causes is often the number of hand-offs required to complete the work. In this video, we'll discuss why it's critical to minimize the number of hand-offs in our value stream. I've listed some work centers or different teams that may be involved when a change is pushed from development to production on this slide. These may be different than those at your organization, but regardless of who is involved, some things are consistent. For example, each handoff requires communication, a request perhaps in the form of opening a ticket, prioritization, scheduling, and testing. With each handoff, some knowledge about the original problem is inevitably lost. In addition, each handoff introduces a potential delay as we wait for someone else to complete the work. The result is the prior scenario. Without escalation, the wait time simply becomes interminable. Consider the relationship between wait time and how busy a resource is. In manufacturing, this resource is often a machine. In IT, the so-called resources are often but not always, people. A major exception is when key services, such as provisioning, test, or production environments require physical hardware. Although I dislike referring to people as resources, I'm using the generic term here because the concept is the same in both manufacturing and technology. As the graph illustrates, wait time skyrockets when resources are more than 80 percent busy. Stated differently, we need to at least 20 percent idle time to avoid long delays. This relationship may seem counter-intuitive. Most of us are busy. At least we wouldn't say that we spent 20 percent of our time twiddling our thumbs and we still managed to get work done without an excessive delay. The issue here is that important jobs must be escalated to prevent them from languishing in queues as the number of handoffs increase. A couple of examples may be helpful. Let's consider the prior list of seven work centers for quality assurance, information security, and IT operations. Assume that each work center operates at 90 percent of its capacity and each job requires an hour on average to complete. In this case, a new job will spend nine hours in the queue at each work center, or 63 hours total across the seven work centers. The resulting lead time is 70 hours, 63 hours waiting, and seven hours working on the job itself. Instead of completing the job in a day, we will have to wait almost two weeks if everyone works normal business hours. There you might think that this example is a bit extreme, but think about the last time that you had to wait a long time in line. My personal favorite is the Department of Motor Vehicles. In my experience, all the customer service agents are busy, and I often wait more than an hour until someone can help me for 5-10 minutes. Why is that the case? It's precisely because all the agents are busy and wait time grows exponentially as their idle time approaches zero. What should we do to mitigate these issues? We need to minimize the number of hand-offs in our value stream if we want to increase flow. Automation is one alternative, including automated tests in self-service portals to create environments. In my prior example, mini Department of Motor Vehicles services, such as renewing a driver's license or vehicle registration, can now be completed online. Even though the work can be done by a customer service agent, I can just as easily type in all the data myself without ever having to leave home. Another alternative is re-organizing our team such that development, quality assurance, information security, and IT operations are integrated rather than organizational silos. Integrated teams can deliver value directly rather than depending on others. Returning to the prior example, in one instance, I had to go to the police station to verify my vehicle identification number before I could register the vehicle at the Department of Motor Vehicles. Well, there's probably a good rationale for having these two government organizations involved. It certainly would have been more convenient and expedient for me if a single organization handled both the vehicle identification number verification and the vehicle registration.