A cost statement tells you what the parts cost. A function list tells you what the product does. Neither tells you where to look. The matrix that joins them does, and it usually surprises the people who designed the product.
Spreading the cost
Components go into the rows with their cost in EUR per unit, functions into the columns. In each cell the team writes what share of that component serves that function. Every row adds up to 100 percent.
Three questions keep the estimates grounded. What is this component for? The pump housing contains pressure (50 percent), guides the flow (20), carries the ports (20) and has extra wall as a corrosion allowance (10). What would disappear if the function disappeared? If the pump needed no corrosion protection at all, the coating, part of the housing wall and part of the material choice would go, and those shares belong to protect product. What does an operation assure? The test run checks flow, tightness, pressure and alignment, so its cost is split over those four functions rather than parked under assembly.
Use steps of ten percent. The matrix is an estimate made by people who know the product, and finer steps add arithmetic, not knowledge. What does add knowledge is the one line of reasoning written next to each split. Three minutes per cell, then move on.
Multiplying each component cost by its percentages and adding the columns gives the cost of every function. For the coolant pump of this series the eight columns add up to the full EUR 220.00 per unit.
Importance from the customer, not from the design office
Cost alone says nothing. A function may be expensive because it is exactly what customers pay for. To find out, the team compares the functions in pairs: of these two, which matters more to the customer? More important scores 2, equally important 1 each, less important 0. Eight functions give 28 pairs and 56 points, and the share of points is the importance of the function.
The answers come from evidence: interviews with key accounts, lost quotations, specifications and two years of service reports. Sales answers first, and anyone who objects has to bring a customer fact, not an opinion. Without that rule the rating becomes a vote on which function was hardest to engineer.
Reading the index
Value index = importance share divided by cost share. Three zones follow.
| Function | Cost EUR | Importance | Value index |
|---|---|---|---|
| Convert energy | 61.50 | 17.9 % | 0.64 |
| Connect system | 28.00 | 5.4 % | 0.42 |
| Protect product | 20.00 | 5.4 % | 0.59 |
| Support rotation | 24.50 | 8.9 % | 0.80 |
| Move coolant | 33.00 | 23.2 % | 1.55 |
| Seal shaft | 23.50 | 19.6 % | 1.84 |
| Enable service | 6.00 | 8.9 % | 3.27 |
Below 0.8 the function costs more than it matters, and that is where ideas are collected first. Between 0.8 and 1.2 cost and importance are in balance: take ideas that are safe, do not force them. Above 1.2 the function matters more than it costs, so it is protected, and the team asks whether spending more there would win orders.
The three search fields of the CP 40 cost EUR 109.50 together, half of the unit cost, and carry 28.6 percent of the importance. Joining the pump to its surroundings costs more than sealing the shaft, a fact nobody had seen, because the cost sits in five different components. The gap of EUR 24.00 will be closed there, not at the impeller.
One warning: the index is a pointer, not a budget. A 1.5 kW motor cannot be bought for EUR 35.00 because a table says so. What the table says is that EUR 26.50 too much sits in converting energy, and the team’s job is to ask which part of that is ours to change.