Buyers always ask about runtime, and for a long time we just quoted the number the LED driver datasheet implied. Then a French account pushed back, saying their customers were getting less. So we sat down and actually measured it properly over two weeks.
The first finding was obvious in hindsight: the headline runtime is the lowest brightness, not the one people use. On high, our 18650 zoom model ran nowhere near the box figure. On the eco low mode it actually beat the spec. Both true, but only one gets printed big.
Second, regulation matters more than the number. A torch that holds steady output then drops off feels honest. One that slowly dims for hours can claim a huge runtime because technically it's still emitting light, just barely. We changed how we quote, giving a regulated runtime and the point where output falls below half.
The cold test was the eye opener. I left two units on the balcony in February overnight, maybe -4°C. The 18650 lost a chunk of capacity and one sample shut off early. We now note a cold-weather derate, because telling a Nordic camper a summer number is just asking for a complaint.
We also caught one of our own batches using a cell that was rated 2600mAh but tested closer to 2300. That got fed straight back to the supplier with our discharge graphs. They weren't thrilled but the data was the data.
End result, our runtime numbers got smaller and our returns got fewer. I'd rather under-promise. A camper who gets more than the box says is a happy camper, and happy campers reorder.
