What is extra processing?
Extra processing, also called over-processing, is doing more work than the customer asked for or needs. A finer finish, a tighter tolerance or one more check can feel like good work. But if the customer does not need it, they will not pay for it, and it costs you time every day.
Quality is meeting what the customer asked for, exactly. More than that is not quality. It is cost.
Where does it hide?
In the shop
- Grinding a surface finer than the drawing asks.
- Holding a tighter tolerance than the drawing needs.
- Deburring edges that will never be seen or touched.
- Painting a surface that will be machined later.
- Checking 100% of the parts where a sample check is enough.
- The same part inspected by two departments.
- Extra packing the customer did not ask for.
In the office
- The same data written in a register, then in Excel, then in the ERP.
- Four signatures on a routine purchase order.
- One report made in three formats for three people.
- A long report, when a one-page summary would do.
What does it cost? An example
A drawing asks for a surface finish of Ra 0.8 on a bearing seat. The operator's habit is to add extra passes to reach a much finer finish, "to be safe".
- Extra time: about 1.5 minutes per piece.
- A batch of 100 pieces: 100 × 1.5 = 150 minutes, which is a third of a shift.
(These numbers are an example.) The customer does not need the finer finish and does not pay for it. Once the drawing requirement was explained and written into the machine's standard work, the extra passes stopped.
Doing exactly what is asked, nothing more and nothing less, is what quality means.
Does extra processing make the product better?
Sometimes people say, "But it improves our product." Two cases are different.
- If it is a deliberate improvement, decided by management, and the customer values it, then it is a choice. Fine.
- If it is extra that nobody asked for and nobody will pay for, it is waste. And it adds more chances for mistakes.
One example: a customer asked for a polished finish, and the plant also gave a matte finish. A job of 8 hours took two to three times longer. Extra packing or paint that the customer did not ask for is the same.
Why does it happen?
- The requirement is unclear. Operators do not know which dimension or finish really matters.
- The habit of "extra safe". Past complaints have made people add margin everywhere.
- Distrust between departments. Each department checks again because it does not trust the previous one.
- Reports and approvals that grew over the years and are never reviewed.
- No written standard. Each person decides how much is enough.
How do we reduce it?
- Read the requirement. Know what the drawing, order and specification really ask for.
- Write it into the standard way of working, so the operator sees the finish, tolerance and checks needed for the job.
- Match the check to the risk. Use a sample check where the process is stable, and a full check only where it is needed.
- Remove duplicate checks. If two departments inspect the same thing, decide who owns it.
- Review approvals. For each routine document, ask: who really needs to sign?
- Enter data once. Choose one place for each piece of information and use it.
- Ask the customer when you are unsure, before you add the extra work.
How to start in your plant
- Choose one product and list every process step and check.
- Next to each step, write which customer requirement it serves.
- Any step with no answer is a candidate for removal.
- Try dropping it on one batch, with the quality person watching.
- If the quality holds, update the standard and tell the team.
Be careful with this review. The goal is to remove only what adds nothing. Never remove a step that protects safety or a real requirement.
Key points
- Extra processing is doing more than the customer asked for or needs.
- More finish, tighter tolerance or extra checks add cost, not quality.
- Unclear requirements and old habits are the usual causes.
- Write the exact requirement into the standard way of working.
- Match checks to risk, and enter data only once.
