China Manufacturing Decoded

Why do some working prototypes still fail when they reach production?

This is episode 329 and the second part of our discussion on this topic, and Adrian and Paul move from the general prototype-to-production gap into real-world failure patterns that can derail a product launch. They look at 3 common scenarios:

Component swaps made for cost reduction

Firmware clean-up before release

And transferring production from one factory to another

You’ll hear why a cheaper component that looks identical on paper can still cause major problems, why every firmware change needs to be tested and documented, and why a factory transfer should never be treated as a simple handover.

The episode also explains how a structured NPI/MPI process, production-representative builds, configuration control, phase gates, pilot runs, and factory process audits help reduce the risk of production failure.

The key message: a prototype proves the concept, but production proves the process. Before approving production, you need to know exactly what was validated, what configuration it applied to, and what has changed since.

 

TIMESTAMPS

00:00 - Introduction: why working prototypes still fail in production

01:32 - Failure pattern 1: component swaps and hidden validation risks

06:26 - Failure pattern 2: firmware tidy-up before production release

08:53 - Failure pattern 3: transferring from prototype shop to production factory

13:20 - How to bridge the prototype-to-production gap

13:48 - Why a structured NPI process matters

14:51 - Production-representative builds, EVT, DVT, tooling, and PVT

16:49 - Controlled ramp-up instead of jumping straight to mass production

17:32 - Configuration control: validation only applies to what was tested

20:29 - Practical decision framework for managers

22:03 - Setting a configuration baseline from DVT onward

23:05 - Using NPI phase gates and change assessment before moving forward

24:29 - Factory process audits: why an audit is not just a factory tour

27:09 - Pro tips: quality standards, NPI discipline, and validation tracking

30:39 - Factory transfers and why pilot runs are essential

33:05 - Final recap: what changed, what was validated, and what is now unknown

 

Related content
Get help with your project from Sofeast. These services cover the topics discussed today:

New Product Introduction Support

NPI Deliverables Review

DFM Review for Manufacturing in Asia

Reliability Engineering & Testing

Process Management Audit (PMA)

First Article Inspection

Get in touch with us Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

A prototype works. The team signs it off. Everyone feels confident.

Then production starts, and unexpected failures appear.

Why does this happen?

In this episode, Adrian is joined by Paul Adams, the Sofeast Group's Head of New Product Development, to discuss the gap between prototype and production. This is part one of a two-part discussion on why working prototypes can still fail once products move toward mass production.

Paul explains why prototypes and production units are often not the same thing, even when they look identical. The episode covers five areas where important changes can creep in:

Components

Firmware

Suppliers and factories

Tolerances and process variation

Validation basis

The key point is simple:

A prototype proves the concept. Production proves the process.

Understanding that difference helps hardware teams, product developers, and importers avoid painful surprises when moving from a successful prototype to production.

In part two, next week, we’ll continue the discussion by looking at common real-world failure patterns, including component swaps, firmware tidy-ups, factory transfers, and how a structured NPI process helps close the gap.

 

TIMESTAMPS

00:00 Introduction: why working prototypes can still fail

02:09 Prototypes and production units are not the same thing

03:46 The gap between prototype and production

04:23 Five things that change before production

04:36 1 - Components: prototype parts vs production parts

09:17 2 - Firmware: why prototype code is not production-ready

12:03 3 - Suppliers and factories: why process knowledge gets lost

16:50 4 - Tolerances and process variation

19:54 5 - Validation basis: What exactly was tested?

22:22 Key takeaway from part one

23:17 What to expect in part two

 

Related content

How Many Prototypes Are Needed Before We Get ‘Perfection?’

Process Management Audit (PMA)

An Effective New Product Development Process for Electronics

From Prototype to Production: 7 Pitfalls for Tech Products

Get in touch with us Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

Host Adrian revisits episode 49 (a ‘gold episode’ originally recorded in 2021), a topic that still catches many product developers and importers by surprise: non-recurring engineering costs, often shortened to NRE costs.

These are the one-time costs needed to get a new product ready for production, such as engineering work, product design, prototyping, tooling, supplier sourcing, reliability testing, compliance testing, testing fixtures, and production setup.

If you underestimate NRE costs, your product plan may look profitable on paper but fall apart before launch. This episode explains what NRE costs are, why they can grow quickly, where they appear in different manufacturing processes, and how to protect yourself with better planning, supplier due diligence, and the right development agreements.

 

TIMESTAMPS

00:00 — Intro: why NRE costs still matter

01:13 — What are non-recurring engineering costs?

03:04 — Why NRE costs affect your real product margin

04:16 — Why NRE budgets often grow during development

07:37 — Typical NRE costs by product and manufacturing process

08:10 — Plastic injection molding and tooling costs

10:44 — Custom PCBAs and electronics engineering costs

13:46 — Why NRE planning affects cost and delivery time

15:53 — Existing tooling, white-label products, and off-the-shelf options

18:51 — IP and dependency risks with ODM products

20:08 — When a manufacturer offers to absorb NRE costs

22:03 — Why a development agreement matters

24:27 — Why manufacturers prefer production over development work

26:39 — A working prototype does not mean you are production-ready

29:04 — Final summary: what to include in your NRE planning

 

Related content

What is an NRE Cost (Non-Recurring Engineering)?

Costs and Milestones to go from Product Concept to Market?

How to Cost Your Product Properly (Design-to-Cost Explained)

Get in touch with us Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

In episode 236, we explore why so many hardware products never make it to market, even when the idea is strong, the team is ready, and the budget is there.

In this episode of China Manufacturing Decoded, your host Adrian is joined by Paul Adams from Agilian, part of the Sofeast Group, to break down the real reasons hardware projects stall before they even start, and what you can do to avoid it.

They go beyond theory and share practical lessons from real projects, including costly mistakes around missing specifications, bad assumptions, and external pressure to move too fast.

You’ll learn:

Why missing product requirements quietly kill projects

The difference between having an idea and being ready to start

How assumptions compound into expensive errors

The hidden risks in BOMs, components, and compliance

Real-world case studies where projects stalled, and why

A practical 10-point checklist to validate your readiness before development

The goal of this episode is to help you avoid delays, wasted budget, and failed launches when you're launching your product.

🎧 Listen now and make sure your next product is built on solid ground.

 

TIMESTAMPS

00:03 — Intro & episode overview

01:01 — The “failure to launch” problem in hardware

02:01 — It’s not the team: real root causes

03:02 — Assumptions & missing information (core issue)

07:00 — Red flags: missing requirements & BOM

11:57 — What “ready to start” actually means

12:45 — NPI process & phase gates explained

14:22 — Specs as a living document (market changes

15:05 — Mechanical, electronics & feature requirements

17:34 — Volume assumptions & pricing impact

19:08 — The danger of rushing decisions

20:44 — Case study: prototyping failure under pressure

24:25 — Case study: component & supply chain risks

26:33 — Case study: regulatory & certification surprises

29:45 — The 10-point pre-start checklist

32:53 — Most common mistake

33:47 — Final takeaway

 

Related content

Transitioning to Manufacturing from Product Development | 2 Options

IP Protection in China when Developing Your New Product [Importer’s Guide]

Bill of Materials (BoM) Explained

Design to Cost (DTC) Explained

Getting To Grips With Non-Recurring Engineering Costs (NRE) [Podcast]

11 Common Electronic Product Certification And Compliance Requirements

Crowdfunding Failures: 4 Great Prototypes That Failed To Launch

Learn more about how we handle DFM & Industrialization (NPI) for our manufacturing customers

Get in touch with us Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

Some product manufacturers treat launch as the end of the journey. It isn’t.

In episode 325 of China Manufacturing Decoded, Adrian and Renaud break down a powerful idea from Tony Fadell: “Builders build, ship, then solve what breaks.”

They explore what really happens after a product hits the market, and why chasing perfection before launch can actually kill your chances of success.

You’ll learn:

Why over-engineering delays launches (and increases risk)

How Version 1 should really be defined: simple, lovable, and complete

What real-world users reveal that prototypes never can

How to collect meaningful feedback without damaging your reputation

Why early adopters are critical for innovative products

How smart teams build Version 2 while Version 1 is still launching

Developing a new product in 2026? You'll have food for thought from this one!

 

Sections

00:13 — Episode overview

00:37 — Tony Fadell’s quote

01:37 — Why perfection is a trap

04:28 — Engineering vs speed trade-off

06:30 — Launch early vs over-engineering

07:46 — De-risking with Version 1

10:30 — “Simple, lovable, complete”

13:43 — Launch isn’t the finish line

15:04 — Real-world user behaviour

17:06 — Nest example (unexpected insights)

19:36 — Managing reviews & early releases

21:27 — Choosing the right early users

24:02 — Misinterpreting “ship early”

25:47 — Lessons from product reliability

26:56 — Why post-launch work matters

28:28 — Continuous product development

30:25 — Key takeaways

Related content

Tony Fadell's LinkedIn post

How to Manufacture a New Product with the Customer Journey in Mind

Buy the book: Build: An Unorthodox Guide to Making Things Worth Making

A Logical Development Roadmap for New Hardware Products

 

Get in touch with us Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB