Source
Executive Summary
- discusses the concepts of flow efficiency and resource efficiency and emphasizes the importance of optimizing flow to achieve real efficiency.
- He explains that resource efficiency, which focuses on keeping people busy, can actually hinder flow by causing context switching and waiting on others.
- Instead, the speaker suggests visualizing the system, stabilizing it through limiting the working process, and making iterative changes to optimize flow.
- He provides examples of how organizations have reduced lead times and improved efficiency by identifying waste, adjusting task order, and creating buffers (slack time).
- By prioritizing flow efficiency, teams can achieve better results and reduce pressure in the work process.
00:00:00
the concept of flow efficiency and how it relates to making an organization more efficient.
- introduces a simple three-stage cycle of change, which includes visualizing, stabilizing, and optimizing.
- Visualizing involves observing and measuring the current processes,
- while stabilizing aims to reduce variability and establish consistency.
- Finally, optimizing allows for improvement and tinkering with the processes.
- also explains how to model a flow or process using a simple diagram, where different people contribute and add value to achieve a desired outcome.
- See note on slack
00:05:00
the concept of lead time and waiting time in a process.
- Lead time is the total time it takes from a request being made to having the final product, and waiting time refers to the time spent waiting for the work to be picked up by the next person in the process.
- gives an example of a trading firm where getting a server took 30 days, but when they analyzed the process, they found that 94% of the time was spent waiting.
- This highlights the inefficiency of many business processes, where a large portion of time is wasted on waiting instead of value-adding work.
- See value stream mapping
00:10:00
Resource Efficiency and why it is not the best approach for maximizing value.
- Resource efficiency focuses on keeping people busy and highly utilized, but this often leads to context switching and waiting on others, which negatively affects flow.
- Dan explains that local optimizations within a process, such as focusing on utilization, can destabilize flow.
- He introduces the idea of cycle time and utilization as two axes in a graph, showing that higher utilization leads to longer cycle times. He emphasizes that resource efficiency is not the most effective way to achieve results and highlights the limitations of this approach.
00:15:00
Flow and how it can be applied to knowledge work.
- example of traffic on a fully utilized motorway to illustrate that when there is no spare capacity, each car’s journey will take infinity time unless the bottleneck is unblocked.
- Similarly, in knowledge work, if there is a long queue building up at testing, it means that testing is over-utilized and the environments are over-utilized.
00:20:00
flow efficiency and how it is different from resource efficiency.
- optimizing flow is the key to real efficiency. resource efficiency is a distraction.
- He suggests visualizing the system through value stream mapping and stabilizing it by limiting the working process.
- As discussed in lean, having too much WIP increases variability of completion. Value stream mapping is looking for average completion time
- He emphasizes the importance of making changes in a scientific and iterative manner to optimize flow.
- this approach can be applied not just in IT or development processes, but in any context where work happens and pieces are connected.
00:25:00
how they were able to significantly reduce the lead time for getting a server from needing it to it being live and ready.
- By visualizing the process and identifying waste, they were able to take a third of the lead time off.
- Then, by making adjustments to the order of tasks and getting decisions made earlier, they were able to take another third off.
- Furthermore, they discovered that they only built three types of servers and created a buffer of each type in each data center, reducing the lead time to just 2 hours.
- This improvement not only made the team look like legends but also reduced pressure and made the work process more relaxed.
Credits
- Summarised using summarize.tech
- Edited by yours truly