Some lanes lost money on every shipment, and looked like growth.

A forwarder handling ocean and air across India, Southeast Asia and the Gulf. Volume had grown for three years and margin had not, which is the classic signature of a mix problem that nobody has decomposed.

Two things were happening at once. Rate requests were answered in arrival order, so the highest-value enquiries waited behind the smallest ones and were lost to whoever quoted first. And lane profitability was reported as a blended average, which meant the routes losing money on every single shipment were being subsidised by the ones that were not, invisibly, forever.

Defining margin honestly

The definition took longer than the build. Finance wanted lane margin after direct costs. Operations pointed out that detention and demurrage exposure was a real cost that fell on some lanes far more than others, and that carrier invoice discrepancies were being absorbed rather than disputed because checking cost more than the errors were thought to be worth.

The signed definition was margin per lane after all recoverable costs, with detention exposure modelled from actual history rather than contractual entitlement, and carrier invoice variance included at its measured rate rather than assumed to be zero.

What changed

Rate requests began to be prioritised by expected margin rather than arrival order. Median response time on business-hours requests fell from 6.5 hours to 2.1. More importantly, the requests that got the fast response were the ones worth winning.

Four lanes were shown to be loss-making at every observed volume. Two were repriced and retained. Two were exited, which cost revenue and improved the P&L, and which the sales team had suspected for two years without the evidence to make the case.