Across 326,000 anonymised claims, the largest driver of loss is not fire, theft or storm — it is damage during loading, unloading and road transit. Loss is not an act of God. It is an operations problem, and operations problems are preventable.
Party names — carriers, shippers, consignees — are coded. Everything else you see here is real: the actual causes, commodities, corridors and seasons that drive loss.
Ranked by value paid, the picture is unambiguous: handling — loading, unloading, jerks and jolts — dwarfs every catastrophic peril. This is the single most important fact for prevention, because handling is controllable in a way that weather is not.
The commodities driving loss are consumer electronics and white goods — displays, refrigerators, air-conditioners, washing machines. High unit value and shock-sensitivity is exactly the profile that turns a jolt into a claim.
A risk model tuned to the average claim would miss the money entirely. The tail is the book.
Most movement is domestic road freight, and loss clusters on specific lanes. A single high-severity corridor can carry more value than thousands of routine intracity claims combined — the signature of concentrated, addressable risk.
The Bengaluru→Mangalore lane is a study in itself: only 181 claims, but a ₹5.2 crore average ticket — a small number of very large losses, exactly where survey and control effort earns its keep.
A carrier is rarely simply good or bad. Coded here as A / B / C, the three highest-loss carriers show sharply different profiles by commodity and lane — the kind of pattern only a pooled, cross-client corpus can reveal, and the raw material for a carrier-specific risk rating.
| Carrier (coded) | On displays | On refrigerators | On washing machines | Worst corridor |
|---|---|---|---|---|
| Carrier A | ₹1,211 Cr 60,382 claims | ₹28 Cr | ₹51 Cr | Bengaluru→Mangalore ₹934 Cr |
| Carrier B | ₹308 Cr 750 claims — high severity | — | — | Bengaluru→Vijayawada ₹307 Cr |
| Carrier C | ₹86 Cr 28,883 claims — high frequency | — | — | Noida→Noida ₹19 Cr |
Carrier B and Carrier C tell opposite stories on the same commodity: B has few claims but enormous severity; C has tens of thousands of small ones. The same rating would be wrong for both — which is why the intersection, not the carrier alone, is the unit of risk.
78% of loss is handling and shock — loading supervision, lashing, packaging and shock-monitoring beat catastrophe cover. The controllable causes are the expensive ones.
Displays and white goods dominate the book. Fragility-graded packaging warranties and pre-dispatch condition capture target the value where it actually sits.
Carrier-alone ratings are blunt. The signal lives in carrier × commodity × corridor cells — and a pooled cross-client corpus already holds them at volume.
68% of loss is the largest 1% of claims, concentrated on a few high-severity corridors. Survey and control effort aimed there returns more than any broad SOP.