The report showed something very simple: the customer had not received their full order. Incomplete. If you look at it only through the KPI, the first thought comes by itself: the driver forgot a bag during delivery.
But when you check the physical process, the picture changes.
In one of the cases we documented, a bag from one section of the warehouse was missing. We also had cases where the bags showed as scanned, but the wrong bag had been picked up or delivered. In other cases, the warehouse turned up bags without a label, bags with two names, mixed trolleys, and even a full trolley missing from the normal loading flow.
The dashboard said: incomplete delivery. The floor made you ask something else. Was it really the driver's mistake at the customer's door, or had the order left the warehouse incomplete?
That is the whole difference between a table and an operation.
The table can tell you: incomplete = 1. It does not tell you that physically this may have happened: the bag was prepared, it entered the process, maybe it was even scanned, but it was put in the wrong trolley or loaded onto the wrong route. In Excel this ends up as a number. On the ramp it is a physical object in the wrong place. That is why the real cause is not always found on the dashboard.
When I wanted to understand an operational problem, I did not stay with the table. I went to the place where the process was happening. For me, that place was mainly the ramp and the loading area.
There you saw things that do not exist in Excel. Whether the trolleys were organised by route. Whether there were mixed trolleys. Whether bags were missing. Whether the labels were correct. Whether the drivers were scanning as they should. Whether dry, chilled and frozen were following the right process. Whether equipment or trolleys were blocking the area. Whether the gates and the parking and loading area were organised. And above all, whether a problem that would later show up as a KPI had started right there.
For chilled and frozen there were also concrete rules. Chilled products had to be scanned inside the cold area, and frozen inside its own area. If a bag was missing, the problem had to be reported, and the trolley was not to be taken out as if everything was fine. A rule like that looks small on paper. On the ramp it is the difference between a problem caught in time and a customer who receives an incomplete order.
We also had ramp audits and direct checks of the process. But the floor is not only a place to look. It is a place to reconstruct what happened.
When a bag was missing, the questions were simple. Where was the bag last seen? Was it in the trolley? Had it been scanned? Was something missing before departure? Who noticed the problem? Were the route and the trolley correct? Had there been waiting or a problem during loading?
With delays it was the same. We also checked the reasons the drivers gave. There were cases where a driver reported traffic, parking or a traffic light, and we had to see whether that matched reality. In some cases we had photos or screenshots of the situation on the road.
What I learned is that a single version is not enough, neither the table's nor one person's. You have to connect what the driver or the dispatcher says with what really happened in the process.
And what you see has to be written down, otherwise tomorrow it is only a memory. For the operation I wrote structured daily reports, where the cases were added one after another in the same structure. For specific cases we used written operational communication, with a description of the incident and its cause. We used photos and screenshots when they were needed to document real situations: traffic or parking on the road, but also physical problems on the ramp, such as disorganised trolleys, boxes or trolleys in the wrong place, open gates or a badly organised loading area.
Over time the checks became much more structured: pre-scanning, matching each trolley to its route, final completeness checks, checks of dry, chilled and frozen, ramp audits and analysis with Pareto and 5 Why. Later I also worked with more organised structures, where every case had the incident, the action, the cause and the KPI follow-up. None of this was born in Excel. It was born from what could be seen on the ramp.
This does not mean the table has no value. Without it you would not even know you had to go to the ramp. But the table is the starting point, not the answer. When you treat the number as the answer, you punish the driver for a bag that was never in his trolley.
In Excel, a mistake is a number. On the ramp, it has an address.