Cold chain is a promise, not a truck
Fresh is a claim you make every single morning, and it is easy to quietly stop keeping.
By Sovandarapor (James) Kong · · Ventures
A crate sits on a pavement while the driver finds the right person at a café. It is thirty-four degrees by nine in the morning here, and fresh milk has been in a slow argument with the climate from the moment it left the animal.
Everyone understands the climate and buys a refrigerated truck. The truck is almost never the problem.
The failures are in the gaps. Product staged on a loading area at ambient temperature while the paperwork is sorted out. A café fridge that is technically a fridge but is opened forty times an hour and has never been checked with a thermometer. A last delivery on a route that has been in the vehicle for six hours, with the door opened at every stop before it.
Each gap is short, and none of them feels significant to the person creating it. Together they decide whether a product reaches its stated shelf life. The customer experiences it as milk that went off early, once, and remembers.
Farms measure at the farm. Retailers measure at the shelf. Both numbers are usually fine. The damage happens in between, across a dozen handovers that mostly happen when nobody is watching, because the vehicle in between belongs to a schedule rather than a department.
The cheapest useful thing we did was record temperature at handover, at the point where responsibility moves from one person to another. Only at the handovers. Then we could see which leg of which route was the weak one, and it was never the leg anyone predicted.
Routes get planned for distance: shortest path, fewest kilometres, least fuel. But the last customer on a route receives product that has been through every door opening on that route. If the last stop is the most quality-sensitive account, the problem is built into the schedule. Ordering the route by product vulnerability and account sensitivity first and distance second costs slightly more fuel, and it is the cheapest quality improvement available.
The most valuable data in a fresh business is what comes back, and the pattern in it: which product, which route, which day, which customer, which reason. That only exists if returns are recorded as their own event. The temptation to adjust the delivered quantity and move on is enormous, because it makes the paperwork simpler and the numbers tidier. It also destroys the only signal there is. What left, left. What came back, came back.
Any producer can print the word fresh. It costs nothing. The version worth something is narrower: this product was at this temperature at every handover, it reached you within this window, and here is what we do when it does not. Every fresh business has failures. In the ones worth buying from, a failure produces a replacement and a recorded reason. In the rest it produces an argument about whose fridge it was.