The real big gaps — the ones even China’s industrial speed can’t touch — are in industrial components that are big, bulky, complex, and high-value. For everything else — the tiny, consumer-speed products with fast iterations (90-day product cycles) — you can’t beat what China offers.
The one-line rule: don’t play Shenzhen’s game — use it. China’s real edge isn’t cheap labor anymore; it’s industrial learning speed and agglomeration (suppliers, talent, and tooling clustered in a 50-mile radius that compounds every iteration).
When to choose China
- You need speed to first sample — sketch to prototype in a week, not a quarter.
- High volume, fast product cycles, tight cost targets.
- Commodity or standard components where the part isn’t your moat.
- You need a dense supplier ecosystem nearby (tooling, PCBs, injection molding all within reach).
- You’re iterating hardware rapidly and want the whole system compounding for you.
When NOT to choose China
- The item is bulky or heavy — freight destroys the economics.
- You need it fast and local: small custom runs, “a weird size by next week.”
- Your value lives in proprietary IP, algorithms, or data — not the physical part.
- Being close to your customer matters more than unit cost.
- The component is existential and you can’t afford a single point of failure across an ocean.
- Reliability, customization, and control beat rock-bottom price for your customer.
China — advantages
- Industrial learning speed: gets better every iteration.
- Agglomeration: entire supply chain within a 50-mile radius.
- Fast sampling and prototyping cycles.
- Low unit cost at scale.
- Deep, mature tooling and manufacturing talent.
China — disadvantages
- Long lead times (often ~8 weeks) plus high minimum order quantities.
- High copy risk — weak IP protection.
- Bulky-item freight kills the math.
- Poor fit for small, custom, fast-turnaround orders.
- Distance from your customer and your market.
- Dependency risk: when one part controls your fate, you don’t really own your product.
Where this leaves us
Our moat was never the manufacturing — it’s the insight wrapped around it: understanding the athlete, owning the algorithm, staying close to the user. Use China for what it’s unbeatable at (speed, density, cost at volume). Keep in-house or nearby what’s existential, defensible, or customer-facing. And audit your dependencies now — name the 3–5 parts that would end you if you lost them tomorrow, then dual-source before you’re forced to.