Component sourcing: catalogue parts against parts made to your drawing

The interchangeability question decides everything downstream

Ask whether a second supplier could deliver a functionally identical part next month without tooling, approval or design change. If the honest answer is yes, the component is a commodity and effort belongs in price, availability and obsolescence monitoring. If the answer is no, you have a captive item and the effort belongs in supplier capability, part approval, tooling arrangements and an exit plan. Engineers routinely create captive parts without meaning to, by adding a small dimensional or finish requirement that eliminates every standard offering, and nobody notices until a second source is needed.

Bespoke parts inherit the process that makes them

A part drawn without reference to a manufacturing route arrives quoted at a price that surprises everyone. Wall thickness that forces a slow moulding cycle, a tolerance that pushes a turned feature into a grinding operation, a radius that needs an additional tool, a surface callout that adds a whole process step: each is invisible on the drawing and dominant in the quote. The productive move is to put a candidate supplier's process engineer in front of the design before it is released, and to accept that the resulting changes are the least expensive cost reduction available at any later stage.

Catalogue parts fail differently, and mostly through change

Standard components rarely fail on capability. They fail through end-of-life notices, silent internal changes by the manufacturer, allocation during shortage, and distributor stock that turns out to be someone else's returned inventory. The controls are different in kind: subscribing to change and discontinuation notifications, buying through channels that maintain traceability to the manufacturer, holding an approved alternate on the drawing where the design allows, and reviewing the component list for lifecycle status rather than only for price. Lifecycle status deserves a place beside price in any component review, since a part that becomes unobtainable mid-programme costs far more than one bought slightly above market. Distributor authorisation is worth checking on anything electronic, because the shortage years taught every plant what unauthorised channel stock can contain.

What the second source really costs on a captive part

Adding an alternative supplier for a drawn component is not a purchasing exercise. It means new tooling or a tooling transfer, a fresh capability assessment, sample submission, part approval, possibly a customer notification if the end market requires one, and a period where two variants exist in the plant. Plants that discover this cost only in a crisis end up staying with a failing supplier because leaving is worse. Decide early, on the small number of components where the exposure justifies it, and carry the cost while nothing is on fire.

Who holds the drawing, and who holds the knowledge

A drawing describes the part; it does not describe how to make it. Suppliers accumulate process knowledge that never appears in your records: the gate position that fills the part, the fixture that holds it without distortion, the tool path that avoids chatter, the deburring sequence that passes inspection. Where that knowledge lives only in the supplier's head, the drawing you own is a weaker asset than it appears. Capturing at least the process parameters that affect the characteristics you care about, as part of part approval, narrows the gap without demanding the supplier hand over its craft.

Frequently asked questions

When should a design use a standard component instead of a custom part?

Whenever the standard item meets the functional requirement without forcing awkward compromises elsewhere in the assembly. The custom part usually looks better on paper because it is optimised, and worse in practice because it brings tooling, approval, a single source and a slow response when demand changes. Reserve bespoke components for places where they earn their keep: genuine performance need, a meaningful assembly-cost reduction, or a feature that differentiates the product to the buyer.

How do you avoid accidentally creating a single-source component?

Review new drawings against what the market actually supplies, not against what is theoretically possible. Flag callouts that only one process or one manufacturer can satisfy, question tolerances tighter than the function requires, and check that named materials and finishes have more than one producer. The most common trap is a tolerance stack resolved by tightening one component rather than revising the assembly approach, which quietly moves that part into specialist territory.

Is it better to buy components through a distributor or direct from the maker?

Direct usually gives a better price at volume, closer technical access and clearer traceability, but comes with minimum quantities and less flexibility. Distributors break bulk, hold stock, handle small releases and can consolidate many lines onto one invoice. Most plants run both: direct for the high-value or technically demanding components, distribution for the tail. The pitfall is drifting into distribution for a critical part and losing sight of who actually manufactured it.