What is continuous improvement? Follow the work after the fix
Understand continuous improvement through a recurring picking problem. See how evidence, a small trial, operating ownership and follow-up fit together.
Every Friday, a warehouse supervisor asks people to be more careful with labels. Every Monday, the same picking errors appear in the report.
The reminder is an action. The team still has no explanation for why the errors keep happening.
Continuous improvement is the ongoing work of making a product, service or process perform better, checking the result and building on what was learned. ASQ includes both gradual changes and larger improvements. It is not limited to saving a few seconds at a time.
For this fictional warehouse, the job is to understand one recurring error and leave the team with a method it can maintain.
Make the problem small enough to investigate
The team finds 12 wrong-size picks among 400 order lines during a week. That is 3%. The count includes every checked line once, whether it contains one unit or several.
It separates product, location, shift and type of error. Nine of the 12 wrong-size picks involve neighbouring bins for two similar products. That is a lead to investigate, not proof that the bins caused all nine errors.
An operator walks through the pick with the supervisor. They compare the display, the label and the physical item. The size difference is visible on the product but hard to read on the shelf.
The team needs careful observation and a consistent counting rule so next week's report measures the same error. It can start that work without setting up a large programme.
Choose a change that tests the explanation
The team proposes separating the two bins and making the size field prominent. It checks storage and safety requirements, updates the location record and tests the display before moving stock.
For the trial, the supervisor records wrong-size picks, all checked lines and additional walking time. Faster error correction would not compensate for a new unsafe route or a stock-location mismatch.
During the trial, four of 500 checked lines have the error: 0.8%. The team also confirms that the new location appears correctly in the system. The lower rate is encouraging. Product mix, workload and other changes still need checking before claiming a lasting gain.
The PDCA example follows the trial logic in more detail. It is useful when you need to compare a prediction with both the result you wanted and something that could get worse.
Use tools where they answer a question
| What the team does not yet know | A useful next step |
|---|---|
| Where requests or products wait | Observe the work and map the flow |
| Which problem accounts for the most lost time or cost | Use a defined measure and a Pareto analysis |
| Why an error occurs | Investigate evidence and competing explanations |
| Whether a proposed change helps | Run a bounded trial with a comparison and checks |
| Whether the gain persists | Keep an operating measure and a response rule |
The 5 Whys guide helps explore an explanation. A DMAIC project gives more structure when variation, measurement and several possible causes need sustained investigation.
A fishbone diagram can collect possible explanations. The team still needs to check them against the work before choosing a fix.
Someone has to own the new method
The supervisor agrees who maintains bin labels, who updates location records and what happens when a new product needs space. New starters practise the revised pick. A later review checks the same error rate and the awkward cases.
If the rate rises again, the team uses the new records to investigate. It does not simply repeat the label project or assume people stopped caring.
Did the improvement hold? follows that period after implementation. Vrolen's PDCA guide applies the same learning cycle to a production example.
Principles and labels need a practical meaning
Continuous and continual improvement are often used interchangeably. Kaizen is closely related, but Lean and Six Sigma describe more than alternative names for this one activity. The Kaizen, Lean and Six Sigma comparison explains how they fit.
Lists of five principles or three Cs vary by author. Lean's five principles, for example, come from Womack and Jones's account, explained in the comparison above. Whichever shorthand a team uses, its review needs to show the problem, observations, trial and follow-up.
For an operating review, keep a short record of the problem, evidence, change, result and next check. Choose fewer changes if that is what the team can test and maintain. A target for the number of initiatives tells you little about whether Monday's errors have stopped.
Sources and example notes
- ASQ: Continuous improvement, ongoing improvement and the use of repeating trials.
- Lean Enterprise Institute: Kaizen, participation, standard work and follow-up beyond an event.
The warehouse and trial are fictional. The rates are 12/400 and 4/500; the comparison alone cannot separate the bin change from other differences between the weeks.