I Work FOR YOU, Not Factories.

I’m Leon Xu, based in Shenzhen with 15 years in consumer electronics — hardware, embedded systems, supply chain, and production. This article is about a structural mistake that quietly costs hardware teams more than almost anything else: treating engineering, component sourcing, and manufacturing as separate, sequential steps.

PCB Design, Components and Manufacturing: Why They Should Not Be Managed Separately

By Leon Xu | Easelink Tech | Shenzhen, China

The Traditional Handoff Problem

Most hardware projects are organized as a relay race. Engineering designs the board and hands it off. Sourcing buys the components and hands them off. Manufacturing builds the product and hands it off. Each stage does its job in isolation and passes the result to the next.

This looks efficient. It’s actually where projects break. Every handoff is a point where context is lost, assumptions go unexamined, and problems that should have been caught upstream get baked into the design downstream. By the time the factory sees the board, the expensive mistakes have already been made.

The alternative is to treat engineering, components, and manufacturing as one connected decision chain — not three separate stages.

A Technically Correct Design Can Still Fail Commercially

Here’s the core problem in one sentence: a design can be electrically correct and still create a commercially difficult product.

The design can be perfect on the bench — every circuit works, every spec is met — and still be a commercial failure, because the parts are too expensive, or unavailable, or impossible to assemble at scale. Electrical correctness is necessary but not sufficient. Commercial viability depends on factors the schematic doesn’t capture: component cost and availability, lead times, manufacturability, yield, and production risk.

A team that optimizes only for electrical correctness is solving half the problem. The other half lives in the supply chain and the factory floor — and it has to be considered at the same time, not after the design is frozen.

Component Availability Should Influence Design Decisions Before the BOM Is Frozen

The single most expensive mistake in hardware development is this: the engineer selects components from datasheets, the BOM gets frozen, and only then does someone check what’s actually available. By that point, the design is locked to parts that are on allocation, obsolete, or unaffordable — and changing them means a redesign.

Component availability should influence design decisions from the start. An engineer who knows the component landscape — what’s available in China, what’s cheap, what’s about to go end-of-life, what has a pin-compatible alternative — designs a product that can be built. This is why BOM auditing belongs at the front of the process, not the back.

The component and manufacturing realities are not a downstream concern. They’re an upstream design constraint, and treating them as such is the difference between a design that ships and a design that gets reworked.

Alternatives and Substitutions Are Not Merely Purchasing Decisions

When a component is unavailable, the response is often treated as a purchasing task: find a cheaper substitute. But substitution is fundamentally an engineering decision, not a purchasing one. A pin-compatible part is not automatically an electrically equivalent part, and swapping one in without validation is how field failures are born.

The right way to handle a component problem is a joint decision: engineering validates the alternative, sourcing confirms availability and cost, and manufacturing confirms it can be assembled and tested. When these three functions operate independently, substitutions get made by whoever is under the most pressure — usually purchasing — with consequences nobody fully owns. The connected model keeps substitution an engineering decision with commercial inputs, not a commercial decision with engineering blind spots.

DFM: Manufacturing Problems That Begin Upstream in Engineering

Many manufacturing problems don’t actually originate in the factory. They originate in the design, and the factory just gets blamed for discovering them. A component placed too close to another, a tolerance that can’t be held, a test point that doesn’t exist — these are DFM (design for manufacturability) issues, and they’re created at the engineering stage.

When engineering and manufacturing are separated, DFM becomes a series of surprises on the production line. When they’re connected, DFM happens during design, where it’s cheap to fix. The difference is enormous, and it’s exactly the kind of technical maturity I describe in evaluating engineering and R&D capability.

How PCB Assembly and Testing Requirements Affect the Design Stage

Assembly and testing requirements reach backward into design. If a board is hard to assemble, the design wasn’t done with assembly in mind. If a board can’t be tested efficiently, the design wasn’t done with testing in mind. Both are design problems that surface in production.

A board designed with assembly and testing in view has the right footprints, the right test points, the right access for probes, the right panelization. A board designed purely for electrical function doesn’t — and the cost shows up as slow assembly, low test coverage, and defects that escape to the field. This is why the PCB engineer’s DFM awareness matters as much as their routing skill.

What Happens When Engineering, Sourcing and Manufacturing Work Independently

When these three functions work independently, a predictable pattern emerges:

Engineering designs for electrical correctness, ignoring availability and manufacturability.
Sourcing buys the cheapest available parts, ignoring electrical and manufacturing implications.
Manufacturing builds what it’s given, discovering the problems both upstream functions created.
Nobody owns the interface, so problems fall through the cracks between stages — and get caught in the field, where they’re most expensive.

The result is a project full of rework, delays, and quality problems that no single team caused but no single team could have prevented alone. The fix is not more effort at each stage — it’s connecting the stages so decisions are made with the whole picture in view.

A Better Coordination Model for Overseas Hardware Projects

The better model treats engineering, components, and manufacturing as a single system, coordinated by someone who can see across all three. In practice, that means:

Component decisions are made during design, not after the BOM is frozen.
DFM and testing are designed in, not bolted on at the factory.
Substitutions are engineering-validated, not purchasing-driven.
One point of coordination holds the interface between engineering, sourcing, and manufacturing — so nothing falls through the cracks.

For an overseas team, this is where a local coordination layer creates real value: it’s the difference between managing three disconnected China-side specialists and having one thread that connects the engineering, the components, and the factory. This is the theme of the operating model article in this series.

Conclusion

A design can be electrically correct and still create a commercially difficult product. Engineering decisions affect component availability, cost, lead times, manufacturability, and production risk — and the strongest hardware projects connect those decisions instead of handing them off sequentially.

The fix is structural: treat engineering, components, and manufacturing as one decision chain, coordinated by someone who can see across it. Do that, and the expensive problems that normally surface at the factory — or in the field — get caught upstream, where they’re cheap.