A value stream comprises all the steps required to deliver value to a customer. Because value streams are cross-functional, spanning many teams and even departments in a large organization, rarely do those involved have a complete understanding of the entire workflow. This video introduces value stream mapping, a tool to visualize complex work systems, so an organization can deliver greater value more efficiently. Many times we think we know what the problem is in our workplace, such as unreliable software supplied by third-party vendors, lack of environments for testing, unrealistic security requirements, or too few staff to handle increasing workloads. Unfortunately, the solutions we devise to these problems rarely consider the hundreds or thousands of steps in a complex value stream, and our solutions often results in sub-optimization as we improve one process without regard for how that process affects the rest of the system. A value stream map provides a holistic view of how workflows through an entire system, supporting strategic improvements across the end-to-end workflow. Let's consider a real-world example of value stream mapping. The cosmetics business office at Nordstrom relied on a cobol mainframe application that allowed managers to associate salespeople with specific product lines, track sales commissions, etc. For years, discussions at annual planning meetings highlighted the need to replace this application, but other priorities always took precedence. Eventually, Nordstrom conducted a workshop to map this value stream, an exercise that led to several surprising discoveries. First, managers lacked data required when completing forms, such as a salesperson's employee number. Second, forms could only be completed on a computer in a back-office, which required managers to leave the store floor. Participants at the workshop conducted several experiments, including removing the employee number from the forms. These experiments reduced the overall processing time by four days. Later the team launched an iPad application, so that managers didn't have to leave the store floor, which reduced the total processing time to seconds. These improvements eliminated the desire to migrate the application off the mainframe, which was originally assumed to be the cause of ongoing issues. Value streams and organizations with more than a handful of employees aren't trivial. This slide lists some of the teams involved in a technology value stream, including development, information security, and operations, as well as individuals with specific roles, such as the product owner and release manager. Mapping the value stream requires engaging with all of these participants to understand their specific responsibilities. A good way to map a value stream is a workshop, with individuals who are involved in the value stream and have the authority to change their portion of it. The goal of the workshop is to identify the major processes in the value stream, in places that limit the fast flow of high-quality work. Each process is annotated with the lead time, processing time, and percentage of work that is complete and accurate for downstream consumers. Creating a value stream map with 5-15 processes should only require a few hours. Let's consider the illustration of a technology value stream. Notice all the processes required to move a customer request to deployment. Also notice the lead time, processing time, and percentage of work that is complete and accurate. Percent C/A in the diagram for each process. In this example, the total lead time is nine-and-a-half weeks, compared to the total processing time of just seven days. Worse, the percentage of work that is complete and accurate is approximately five percent for the entire value stream. These realities can be sobering, particularly when people realize how dependent they are on processes outside their control. For example, waiting for a sign off from a change approval board delays deployment by one week, which doubles the time between the sprint review and customer availability. The value stream map maybe the first time that participants in the workshop have seen the entire value stream. Participants are often surprised by the amount of work and heroic individual efforts required to deliver value to customers. For example, operators may realize the impact of developers waiting for environments for testing, and developers may realize the extensive effort required to deploy code into production, long after they've marked a feature as complete. Armed with this data, we can focus on the metric we want to improve, be it lead time or the percentage of work that is complete and accurate. The next step is to create an idealized value stream map, which identifies the goals for improvements. Leaders not only help set the goals, but also emphasize learning and problem-solving to achieve the desired outcomes. In a nutshell, leaders empower teams to test hypotheses regarding how to solve a problem, perform experiments to test those hypotheses, and interpret the results. This cycle is then repeated, gradually bringing each team closer to its objective until reality matches the idealized value stream map.
