Collecting ideas is the part everyone wants to start with, and it is the part that fails when the first three steps are skipped. Ideas gathered without a search field repeat the last cost reduction round: the same suggestions, the same objections, the same two percent. Ideas gathered per function, starting with the functions that cost more than they matter, go somewhere else.
Six questions per function
Take one search field at a time. Put the function and its cost on the wall and work through six questions.
Can the function be dropped or merged? On the coolant pump the base plate existed because the first version in 1998 had a motor that could not carry itself. The feet can be cast onto the bearing bracket. Can the function be reduced to what is required? The bearings are sized for 40,000 hours, the specification asks for 20,000. Is there a standard solution? A special flange motor competes with a standard frame motor and a standard coupling. Can one part serve two functions? A bracket can support the rotation and stand on the floor. Can variants be reduced? Two housing variants, one with threaded and one with flanged ports, became one housing with threads and a flange kit as an option. What do others do? Two of three competitor pumps on the table wore a single powder coat instead of two coats of wet paint.
The twelve ideas of the CP 40 came from exactly these sources: seven from the search fields, two from requirements above the need, two from competitor products, one from a service technician.
Rules for the meeting
Twenty minutes per search field. Five minutes of silent writing before any discussion, because an open round gives the floor to the fastest talker and loses the rest. No rating in the room: “we tried that in 2019” and “the customer will never accept it” are written on the card as a risk and dealt with later. For purchased parts above five percent of the cost, invite the supplier, who usually knows better than anyone what makes the part expensive.
Separating collection from rating is not a nicety. It is what keeps the eleventh idea alive long enough to be checked, and the eleventh idea is often the one that works.
Four numbers and one gate
Every idea that survives the clustering gets four numbers, agreed between design, purchasing and controlling: saving per unit in EUR at full cost, one-time cost for tools, tests and documents, the payback that follows from both at the annual quantity, and a risk rating. Each number has an owner who confirms it within two weeks.
Then comes the gate, and the gate is the customer. An idea with a minus on a function whose value index is above 1.2 is rejected, whatever it saves.
| Idea | Function | Decision | EUR per unit |
|---|---|---|---|
| Standard frame motor and coupling | Convert energy | Go | 7.00 |
| Feet on the bracket, no base plate | Connect system | Go | 4.50 |
| One powder coat instead of two wet coats | Protect product | Go | 3.20 |
| Impeller in reinforced plastic | Move coolant | Reject | 6.00 |
| Lower cost seal from a second source | Seal shaft | Reject | 4.00 |
| Seal change from the motor side | Enable service | Go | -0.60 |
The two rejected ideas would have saved EUR 10.00 per unit. Nine of fourteen key accounts specify a metal impeller because of chips in the coolant, and leaks are the most frequent complaint in the service reports. Both ideas attack what customers value most. One idea goes the other way: changing the seal from the motor side adds EUR 0.60 and cuts the seal change from 45 to 20 minutes on a function with an index of 3.27.
Nine ideas were decided, with a net saving of EUR 24.80 per unit against a gap of EUR 24.00, and a one-time cost of EUR 82,000. Before any of them is released, each one is checked against the requirements it touches.