Lead time
13.3 days (8 hours)
Of which waiting
13 days (8 hours)
Flow efficiency
1.9%
process time as a share of the lead time
Right first time, whole process
7.0%
all steps in a row
Process steps with data box
Start with the customer and work back; process time in minutes, wait time before each step in days.
Kaizen bursts: where we will investigate
- Complete the file: Missing documents: which ones, and why are they only discovered here?
- Make the calculation: Only 1 in 6 right first time: understaffing, turnover, no work instruction.
Take each burst to the Ishikawa diagram or 5 Whys to find the root cause.
The map is done, with 2 bursts.
Have the map validated by colleagues from the process who were not in the room, then investigate the bursts. After that: draw the future state without the waste.
- Flow efficiency 1.9%: only 1.9% of the lead time is processing, the rest is waiting. That is normal for an office process, and it means working faster helps little. Look at the queues.
- Only 7.0% gets through the whole process right first time. Every rework adds to the lead time; quality is the speed problem here.
- Lead time 13.3 days against a target of 10 days. The difference is 13 days of waiting; that is where the gain is.
What a Value Stream Map shows
A VSM is a process map with data. Not just which steps there are, but how long each step takes, how long the work sits waiting before it and how often it goes right first time. That makes waste visible: inventory, wait time and rework. The name "value stream" is misleading: the map contains all the steps, including those that later turn out to add no value.
How to build one
- Start with the customer. You draw a VSM from right to left: first what the customer receives, then the step before that, and so on. Name the main supplier as well.
- Always with the people from the process, on brown paper with sticky notes. Afterwards you put the result neatly into the tool.
- Fill in the data box for every step. Process time (how long someone actually works on it), wait time before the step (how long the work sits) and right first time. Measuring beats estimating; estimating beats nothing.
- Read the timeline. Waiting high, processing low. In an office process the flow efficiency is often below 10%: working faster helps little, the queues are the problem.
- Put kaizen bursts where the data points: the longest wait time, the lowest quality. That is where you investigate in the Analyse phase.
Example from practice
In our e-book a Black Belt maps the process of a pensions department where customers complained about errors. For every process step the team measured how often a file was right first time. In some places that was one in six. That is where the kaizen bursts went, and the investigation that followed produced the root causes: understaffing, a focus on speed, staff turnover, and no proper work instructions.
More background is in the knowledge-base article on the Value Stream Map.