In this essay
1.06%. The target was 0.8%.
That is how Damage looked in the April 2025 report. In last mile it was one of the main KPIs we followed every day, together with Incomplete, Delay and Loading Time.
The number was right. But it did not say whether the damage had started on the cart, during loading, in the van or at the customer's door. It only said that we had a problem.
This essay opens a new series, “Improvement that holds”. I want to open it honestly, because an improvement that holds starts from what you really know, not from what sounds good in a presentation.
What I have and what I do not
I have two figures on record. In April 2025, Damage was 1.06% against a 0.8% target. A June 2025 report showed 0.97%.
That is two reports, not a series. I cannot say whether the change was a trend, the season or chance. And I do not present it as the direct result of what I did.
I do not know precisely where most of the damage happened: in the warehouse, during loading, in the van or at the customer. I do not have a documented case where I could describe the damage, the cause and the corrective action without inventing details. I do not have my first concrete step after taking on the role written down either. I will not fill these gaps with sentences that sound good.
What I do know is how we followed it and what problems we saw.
How we followed it
We followed Damage in the operational reports and in the daily stand-up. We did not look at it on its own. We analysed it alongside the other indicators and the everyday problems of the operation.
The problems we saw were operational: how the load was organised, how the carts were used, how the products were placed. The operation had standards for safety, for organising the ramp, for loading and for handling products. The Shift Leaders and Dispatchers followed the operational problems, and coordinating with the DSPs was part of the daily work.
In Pareto and 5 Why in practice I wrote about the measures that came out of the loading cases in that period. But I do not have an exact record of the concrete changes I put in place, or of how each group's behaviour changed. So I do not write them up here as a success story.
A percentage has no address
This is what has stayed with me most from working with damage.
A KPI tells you that you have a problem. It does not au
A damage rate gathers into one number things that started in different places. A product placed badly on a cart. A load organised in a hurry. A box that moves in the van. A difficult delivery at the customer's door. In the report they all become one line.
And damage is usually reported where it is found, not where it starts. I wrote this in Mistakes get lost between departments too, about mistakes in general. With damage it is even easier to get it wrong, because when a damaged product is found, someone always has it in their hands. That person very easily becomes the answer.
If a manager cannot say at which stage damage starts, they are working on the percentage, not on the process.
Two ways t o lower a number
Here I am giving an opinion, not a case from my own experience.
A percentage can move without anyone understanding the process. You can put pressure on the team where the damage is reported. You can make it the topic every morning. The number may come down for a while. But if the cart is filled the same way, the load is organised the same way and the standard stays just as unclear, the damage will come back as soon as attention goes somewhere else.
The other way is slower. You follow the damage back through the process to the stage where it started, and you change something there.
Managing damage should not only be an effo
That is also what the name of the series means. An improvement that depends on the manager's attention lasts as long as the attention does. An improvement that changes the process holds even when the manager is busy with something else.
What I would do t oday
Today I would focus on finding the stage where the damage starts, not only on the person or team where it is reported.
That does not mean removing individual accountability. If a clear and workable standard was not followed, that has to be said. But that question comes after you understand where the damage happened, not before.
For every damage case, I would look for four answers.
Checking one damage case in four questions
1. Where did the damage happen? The stage of the process, not the person: preparation, the cart, loading, transport, delivery. If the answer is “we do not know”, that is a finding too. It means the process does not record where damage starts, and that is where the work begins.
2. Which standard was not followed, if there was one? If there was no standard for that stage, the problem is not the person. It is the system, which left that person to decide alone.
3. Was the standard clear and workable? Not on paper, but on the shift, with that day's volume, time and people. A standard that can only be followed when there is spare time is not a standard. It is a wish.
4. What has
When these four answers are gathered for many damage cases, something starts to show that the percentage does not: the stages where damage starts most often. Damage Control does exactly that. It logs every damaged unit with the stage of the process and the cause, and shows where to start. You can fill in the card below for a single case and copy it for your team.
The percentage tells you how much. The questions tell you where. And only the “where” can be changed.
The goal is not a lower percentage in the repo
Checking one damage case
0/4