Ask most engineers how they source a part, and you'll get a fast answer: Buy it off the shelf, or build it from scratch. Make or build,. standard or custom — it's a clean mental model, and it's wrong often enough to cost manufacturers real time and real money.
The same pattern happens over and over. A part doesn't fit the standard catalog exactly, so the team assumes it needs to go custom. Nobody's being careless; it's just the fastest decision available when you're staring down a build schedule. But "fastest decision" and "right decision" aren't the same thing, and the gap between them is where a lot of unnecessary cost and complexity quietly accumulates.
When companies treat sourcing as a two-option choice, it forces every part that isn't a perfect catalog match into the custom bucket, whether it belongs there or not. And once something is labeled "custom," it triggers a whole chain of downstream work, including engineering drawings, supplier qualification, tooling and quality documentation – all of which might be appropriate for a genuinely unique part, but are wildly excessive for one that just needs a different bore diameter or is a non-standard length.
Many engineering organizations don’t always have a clear, consistent way to test whether a component actually requires full customization before it gets treated as if it does. The default is often instinct, precedent, or simply "That's how we did it last time." None of those are bad instincts, but none of them are decision frameworks, either.
Going Custom Too Early
The costs of a premature custom decision rarely show up as a single line item, which is part of why they're so easy to miss. They show up as:
- Extra engineering hours, because creating drawings and specs for something that already exists in a configurable form, just waiting to be dimensioned to spec;
- Long supplier-qualification cycles, involving the vetting of a new manufacturing partner, auditing, and running first-article inspections for a part that a pre-qualified configurable supply chain could have delivered already validated;
- Extra lead time, often the most painful one, involving quoting, tooling, and production cycles that are measured in weeks. When that timeline collides with a program schedule, it becomes the thing that pushes the whole product launch, and
- Higher supply-chain risk, because every net-new custom part is a net-new dependency — a new supplier, set of specs to manage, and single point of failure if that supplier hits a snag.
None of this means custom manufacturing is a bad choice. Genuinely unique geometries, tight tolerances or novel materials absolutely require it, and no configurable catalog will ever replace real custom machining for those cases. The problem lies in reaching for custom manufacturing reflexively, before checking whether the part actually demands it.
The Middle Ground
Here is where configurable components earn their place, and where they're most misunderstood. A configurable component is a distinct category that solves a distinct problem: You take a proven, pre-engineered base design and adjust specific parameters — length, diameter, thread, material, finish — to match your exact application.
That distinction matters because it changes the entire cost and time equation. You're not qualifying a new part from zero, or waiting on tooling. You're specifying dimensions on a design whose manufacturing process, quality controls and supply chain are already established. The engineering work that would normally go into a custom drawing gets replaced by a configuration step, and the qualification work that would otherwise go into vetting a new supplier is largely already done.
Configurable components tend to solve a specific set of problems well: those that need to fit a particular envelope or mounting pattern but don't need a proprietary geometry; where the custom element is really just a dimensional variation, not a functional redesign; and where a team needs speed and reliability more than it needs a fully custom solution. In factory automation and machine building especially, an enormous share of what gets treated as custom actually falls into this category once someone stops to check.
Building a Real Evaluation Process
If there’s one change that an engineering or procurement team should make, it‘s to build an actual decision point into the sourcing process before a part gets labeled custom by default. That means asking a few questions early:
- Does this part need a genuinely novel geometry or material, or does it need a dimensional variant of something that already exists?
- What specifically about this requirement can't be met by adjusting parameters on a proven design?
- What's the real cost of the qualification and lead-time work a custom designation triggers, versus the cost of a configurable or standard alternative?
- Who on the team is actually accountable for making this call, and do they have the information to make it well?
That last question matters more than people give it credit for. In a lot of organizations, engineering makes the technical call and procurement inherits the sourcing consequences, and the two groups don't always have a shared framework for evaluating customization level together. Getting that conversation to happen earlier — ideally at the design stage, not after a drawing is already locked — is where the real savings show up.
Why This Decision Outlasts the Part
It's worth remembering that a sourcing decision made in a design review doesn't stay contained to that one part. It ripples into how scalable, resilient and quick-to-market the whole product ends up being
A product built on a foundation of unnecessary custom parts inherits every one of those parts' qualification timelines, supplier dependencies, and lead times, multiplied across a bill of materials. Scale that product up, and you're scaling complexity with it. Hit a supply disruption, and every custom part is a separate point of exposure with its own recovery timeline. Try to compress a launch schedule, and custom lead times become the ceiling on how fast you can actually move.
Manufacturers that get disciplined about this — treating standard, configurable and custom as three genuinely different tools rather than a spectrum from "easy" to "real" — tend to end up with bills of materials that are easier to scale, easier to de-risk, and faster to bring to market. It’s not because they avoided custom manufacturing, but because they reserved it for the parts that actually needed it.
The manufacturing environment right now rewards that kind of discipline. Lead times matter more than they used to. Supply-chain resilience is now a planning input, not an abstract concern. And engineering teams are already stretched thin enough without spending hours re-engineering something that a configurable option could have delivered in days. The middle ground between standard and custom is a strategic option that a surprising number of teams don't fully know they have — and using it well starts with asking the right question before the part ever gets labeled.
Dave Evans is president and chief executive officer of Misumi Americas.
