Stage 1 · See · 03 · 25 min
Find the problems and their causes
You know how the process runs. Now find out what is wrong with it, and why.
This module turns the walk, the data and the as-is model into an issue register: every problem in the custom order process, with its cause, where it happens and how often. Module 5 puts a price on each one.
Apply
1. Classify every activity
What you do here: decide, from the customer’s point of view, which activities add value.
Where does each activity belong?
Machining and assembling the coupling
7 more to place
Value-adding
The customer would pay for it
Business-necessary
Required by the business or by law, the customer does not care
Waste
Adds nothing; exists because of a problem elsewhere
The goal is not a perfect sort. It sets the direction: remove waste, make business-necessary steps as small as possible, protect what adds value.
Apply
2. Name the waste
What you do here: match observations from your walk to the eight types of waste.
Lean names eight wastes, remembered as TIMWOODS: transport, inventory, motion, waiting, overproduction, overprocessing, defects and unused skills. Naming the type helps you find the right fix later.
Which waste is each observation?
Orders wait four days before engineering review
Parts are remade after failing inspection
Drawings are printed and carried to the planning office
Engineers walk to Sales several times per order to clarify specs
Finished parts pile up in front of the weekly inspection
A full engineering review for a changed bore diameter
Detailed engineering estimates for enquiries that rarely become orders
Senior engineers answer status emails for hours a week
Waiting and defects cost the customer most. The others cost Northstar time and money every day.
Apply
3. Find the handoffs that bounce
What you do here: find where work goes back and forth between departments instead of flowing on.
A handoff is where work changes department. A ping-pong is a handoff that bounces: questions, missing information, corrections. You count the messages per order between each pair of departments in a sample of orders.
Average number of back-and-forth messages per custom order between two departments, from a sample of 30 orders.
| Between | Messages per order | Typical content |
|---|---|---|
| Sales and Engineering | 3.2 | Missing dimensions, unclear material, customer changes |
| Engineering and Planning | 0.4 | Drawing release |
| Planning and Purchasing | 1.1 | Material availability |
| Production and Quality | 0.6 | Inspection slot |
| Logistics and Finance | 0.2 | Delivery confirmation |
Which handoff bounces most?
Apply
4. Find the root cause
What you do here: trace a symptom back to a cause the process can fix.
A fishbone diagram sorts possible causes by category before you dig. For late custom orders at Northstar it looks like this:
| Category | Possible causes of late custom orders |
|---|---|
| Method | One queue for every custom order, small variants behind redesigns |
| Information | Orders reach Engineering with missing specs |
| Systems | Order retyped from CRM to ERP; capacity in a private spreadsheet |
| People | Only Anna knows the checks that prevent rework |
Then pick one effect and keep asking why. Here: 15 % of custom orders fail final inspection.
Why do 15 % of custom orders fail final inspection? The inspection reports say:
Apply
5. Build the issue register
What you do here: decide what belongs in the issue register and what does not.
An issue describes a problem and its cause. Workshops produce plenty of statements that look like issues but are solutions in disguise or blame. Neither helps you decide what to change.
Which statements go into the register?
Custom orders wait four days before engineering review
7 more to place
Issue
A problem, with a cause you can act on
Solution in disguise
Jumps to an answer before the problem is clear
Blame
Points at people, not at the process
What is left is Northstar’s issue register for the custom order process:
| Issue | Root cause | How often |
|---|---|---|
| Failed final inspection, part remade | Drawing revisions agreed by phone never reach production | 27 orders a quarter (15 %) |
| Orders shipped late | Every custom order waits in one engineering queue, small variants behind redesigns | 76 orders a quarter (42 %) |
| Order retyped from CRM into ERP | No integration between the two systems | Every order, 180 a quarter |
| Capacity planned in a private spreadsheet | The ERP has no usable capacity view | Every week |
| Customers chase order status | Nobody tells them where their order is | 412 enquiries a quarter |
The “solutions in disguise” are not thrown away. They wait until the register shows which problem they would solve.
Reflect
Check your understanding
Three questions on the module.
Final inspection adds no value for the customer. What do you do with it?
Take away
What to keep
- Classify activities: value-adding, business-necessary, waste.
- Name the waste type: it points to the fix.
- Handoffs that bounce mean information is missing upstream.
- Ask why until you reach a cause the process can fix, never “human error”.
- An issue register holds problems and causes, not solutions or blame.
After this module you can
- Classify activities as value-adding, business-necessary or waste
- Recognise the eight types of waste in a real process
- Find handoffs where work bounces back and forth
- Trace a symptom to a root cause the process can fix
- Build an issue register of problems and causes