Why the cheapest part is the one you deleted

Where the cost of a component actually accumulates.

The cheapest part on any assembly is the one that was never added at all, because a component's purchase price is only the smallest and most visible fraction of its real cost, the rest sitting in the drawing time, the tooling, the unique part number, the inventory it now occupies, the supplier who has to be qualified to make it, and the assembly and inspection labour it demands on every single unit built for as long as the design exists.

What is actually happening

A single new component entering a design triggers a whole chain of costs that never appear on the part's own price tag. Someone has to draw it, someone has to review that drawing, a part number has to be created and entered into every system that tracks inventory, purchasing and assembly instructions, a supplier has to be found and qualified to actually make it reliably, and stock of it has to be held, counted and reordered indefinitely once production begins. None of that overhead scales down for a cheap part, a ten-cent washer generates very nearly the same administrative and inventory burden as a much more expensive machined bracket, which means the purchase price on the drawing is a poor guide to the component's true cost to the organisation carrying it for years afterward.

The single-purpose-kitchen-gadget comparison

A garlic press or an avocado slicer bought cheaply in a moment of enthusiasm feels like a nearly free convenience at the till, and in one narrow sense it genuinely was, but it still needs its own space in an already crowded drawer forever afterward, still has to be found, washed and put away every single time it is used, and still sits there taking up room on the rare weeks nobody reaches for it at all. A good kitchen knife already in the drawer, capable of doing the identical job along with several others, adds none of that ongoing cost, since it was already there and already being maintained regardless, and using it instead means the drawer never gained a new resident at all. Getting rid of the gadget later carries its own small tax too, a moment spent deciding whether it might still be useful someday, another spent finding somewhere for it to actually go, the same hesitation that keeps an engineering organisation carrying an obsolete part number on the books years after the assembly that needed it was last built. A component added to an engineering assembly behaves exactly like the single-purpose gadget rather than the knife, its purchase price trivial and its ongoing presence, in every drawing, every stock count, every assembly instruction, a real and recurring cost for as long as the design exists.

Why the accumulated cost never appears on any single invoice

Nobody ever receives a bill that says "cost of maintaining this part's existence," which is exactly why this category of cost is so easy to overlook when a new component is being justified purely against its purchase price. It shows up instead scattered across a dozen different budgets, a purchasing department's supplier management overhead, a stores department's inventory carrying cost, an engineering department's drawing and documentation time, a production line's incremental assembly seconds multiplied across every unit ever built, each individually small enough to be dismissed and collectively large enough to dominate a part's true lifetime cost many times over its purchase price alone. Nobody signs off on that total because nobody ever sees it added up in one place, which is exactly how a design accumulates far more parts than it actually needs without any single decision along the way looking obviously wrong. Adding a part is a single, visible decision made once by one engineer at one moment in a design review, while removing it later requires convincing purchasing, quality and production simultaneously that it is truly no longer needed anywhere, an asymmetry that means the easy direction is almost always to add and the hard direction is almost always to delete, regardless of which one the true, accumulated cost actually favours.

One figure worth keeping in mind

A component's purchase price is very often only a small fraction of its true cost across the life of a product, once documentation, qualification, inventory carrying cost and the incremental assembly time it adds to every single unit built are all counted honestly, which is why a redesign that deletes a component entirely, rather than sourcing it more cheaply, so often produces a larger saving than any negotiation with a supplier ever could.

Why the bill of materials is where this cost becomes visible

A bill of materials, the complete list of every part and quantity needed to build a product, is where this accumulated cost finally becomes possible to see clearly, since a bill of materials with fewer distinct line items is not only cheaper to build in raw parts, it is genuinely cheaper to manage, since every line represents an ongoing administrative and inventory burden regardless of how little that particular line cost to purchase. Two competing designs quoted at the identical total parts cost can carry very different true costs once their bills of materials are compared side by side, the one with fewer distinct part numbers almost always winning once documentation, purchasing and inventory overhead are added back into the comparison rather than left out of it entirely. A quote that only ever compares the sum of the purchase prices is comparing the smallest and least revealing part of the whole picture. Some organisations track a design's total distinct line-item count as its own metric for exactly this reason, treating a rising part count as an early warning worth investigating on its own, well before any individual part's price looks unusual, since the accumulated overhead this article describes only shows up when the whole list is looked at together rather than one line at a time.

Where this stops being true

A component genuinely earning its place, doing a job nothing else in the assembly can do, is not a cost to be eliminated, it is a cost correctly paid for real function, and this article is not an argument for stripping a design down past what it actually needs. The argument is narrower and sharper than that: a component added for marginal convenience, for habit, or because deleting it would mean solving a slightly harder design problem, is paying this entire accumulated cost for a benefit that a more careful design might never have needed to buy in the first place. Asking, honestly, whether a part earns the ongoing cost this article has described, rather than only whether it earns its own purchase price, is the actual discipline behind every genuine part-count reduction worth making.