Across the engagements we have run, margin erosion does not hide in the same place in every sector. In industrial and manufacturing work it tends to concentrate in unused capacity and complexity cost. In healthcare services it gathers around cost-to-serve and the mismatch between effort and reimbursement. In distribution and logistics it hides in the small order, the long tail, and the customer everyone assumed was fine. What follows is directional, anonymised, and aggregated across many engagements, and it is a sense-check, never a target.
How to read this, and how not to
This is a pattern read, not a benchmark you should hold your own numbers against. We are describing where, in our experience across 30-plus countries, the erosion tends to sit, not how much yours should be. Two firms in the same sector can carry their losses in completely different joints. If you take one thing from this, take the question it points at, not a percentage to chase.
We are deliberately speaking in patterns and directions rather than precise figures. Anything more exact would imply a uniformity the data does not have, and would invite people to treat a composite as a target. It is not. A benchmark used as a target is one of the more reliable ways to make a business worse, because it pushes managers to hit a ratio rather than understand the structure underneath it. The value here is the location of the leak, not its size.
Industrial and manufacturing: capacity and complexity
Two recurring culprits. The first is unused capacity that has been smeared across product cost rather than shown as its own line, which makes a perfectly healthy product look expensive in a slow quarter. The second is complexity: the short run, the special variant, the product kept alive for one customer. These rarely carry their true cost, because the setups, changeovers, and small-batch handling get diluted across the whole catalogue.
The sense-check: look hard at your lowest-volume, highest-variety products, and look separately at how you treat idle capacity. The erosion is usually one of those two, and often both. The honest version of the manufacturing margin question is rarely which product is unprofitable. It is which product was being made to look unprofitable by a cost system that punished it for low volume and complexity it did not control.
Healthcare services: cost-to-serve and the reimbursement gap
Here the margin tends to leak through cost-to-serve that nobody has measured against what each service line actually brings in. Effort and reimbursement drift apart. A service that consumes disproportionate clinical and administrative time can sit alongside one that is comparatively light, and the headline price tells you nothing about which is which.
The sense-check: pull apart effort by service line and set it against reimbursement, not against revenue in aggregate. The painful service is usually the one everyone is proud of and no one has timed. Reputation and reimbursement are not the same thing, and the gap between them is where service-line margin tends to quietly disappear.
Distribution and logistics: the small order and the long tail
The classic shape. A long tail of small orders, small customers, and odd delivery profiles, each individually trivial and collectively a serious drag. Picking, packing, handling, and the cost of a delivery do not scale neatly with order value, so the small order is very often loss-making while looking like activity.
The sense-check: build a customer or order profitability curve, the whale curve, and look at the flat, sinking tail. The customers everyone assumed were fine are frequently the ones sitting in it. They pay on time, they never complain, and they place a steady stream of orders too small to cover the cost of handling them. They are pleasant to deal with, which is precisely why no one looks.
Why does the pattern hold across sectors?
Because the mechanism is the same underneath. Margin hides wherever cost is averaged instead of traced. Spread a cost evenly and you make the heavy users look cheap and the light users look expensive, in every sector. The sectors differ only in which cost is most worth tracing: capacity in the factory, time in the clinic, the order in the warehouse. The fix is always the same instinct: stop averaging, start tracing.
Takeaway: use this as a place to point your attention, not a number to hit. Find the cost your sector is most prone to averaging, then trace it properly and see where the curve sags. If you want to model your own version of this rather than read ours, that is what the platform is for: /costctrl/.