Before approaching a factory, a product team should clarify at least five things: whether the product definition is manufacturable; whether tooling investment, target cost, and order volume are aligned; how sample and mass-production quality will be verified; how changes, risks, and compliance will be managed; and who will inspect before shipment and close the loop on problems.
Many problems that appear to originate at the factory do not begin there. They often start when a team asks suppliers to quote, prototype, or cut tooling before the product has been fully defined.
Is the Product Fully Defined, Manufacturable, and Commercially Viable?
An idea is not yet a product definition. “Build a premium portable speaker” or “build a minimalist keyboard for creators” indicates a direction, but it does not give engineering and manufacturing teams enough information to make consistent decisions.
Before selecting a supplier, the team should answer:
- Who will use this product, in what situations, and what user need will it meet?
- What problem must it solve better than existing alternatives?
- Which features are essential, and which are optional?
- What are the requirements for performance, size, weight, materials, appearance, battery life, software, packaging, and accessories?
- When cost, schedule, appearance, and performance conflict, which trade-offs are acceptable?
These are the foundations of a product definition, and they need to be resolved before a factory starts work. Changes made after development begins consume time, money, and management attention.
Manufacturability also depends on design execution. Industrial design attracts the most attention, but internal mechanical design is often more important at project level. Space utilization, rattles in finished units, drop-test performance, and other validation issues determine whether the product can be built reliably.
Is the definition clear enough to avoid rework once the factory optimizes it for production? Does it address a specific user and use case? Can the commercial model work? Can the proposed industrial and mechanical design actually be tooled and manufactured?
Are Tooling Investment, Target Cost, and Order Volume Aligned?
Tooling cost, unit cost, and order volume are interdependent and should not be approved or evaluated in isolation.
Highly customized products may require molds, fixtures, test equipment, packaging structures, engineering time, and additional certification work. A basic box-shaped enclosure can be inexpensive; a complex design involving two-shot molding or waterproofing can cost ten times as much—or more.
Full tooling makes sense only when projected sales and gross margin can recover the investment. Otherwise, adapt an existing factory platform and validate the market first.
Build a basic profit-and-loss model before requesting quotations:
- Target retail price, landed cost, and ex-factory price
- Initial order quantity and projected lifetime volume
- Maximum affordable tooling and development investment
- Allowances for defects, returns, warranty, freight, duties, and channel costs
Together, these conditions determine whether the product should use dedicated tooling or a lightly customized existing platform.
Have Sample, Mass-Production, and Acceptance Standards Been Defined in Advance?
Quality control should begin before production—not when finished goods are already waiting to ship.
Teams often approve a sample after checking only appearance and basic functions, without documenting the key acceptance conditions. During mass production, the factory may regard a variation as a normal tolerance while the brand treats the same issue as a defect. That is where conflict begins.
Before mass production, both parties should define:
- The final approved reference sample, critical dimensions, and tolerances
- Cosmetic standards for color, texture, gaps, scratches, and printing
- Incoming-quality, functional, performance, reliability, life-cycle, packaging, and drop-test methods
- Firmware and application versions
- Sampling method, sample size, and defect classification
- Conditions requiring rework, replacement, or shipment rejection
Quality control cannot stop at cosmetics. Mechanical and software-related criteria—such as button actuation force and user-interface behavior—must also be defined. Product acceptance needs measurable technical requirements and inspection criteria for the key user experience.
How Will Milestones, Changes, Risks, and Compliance Be Managed?
Every project will change. The real problem is not change itself, but change that escapes control. A component may become unavailable, reliability testing may require a structural revision, marketing may request another feature, or certification testing may expose a new design problem.
Every change can affect cost, schedule, quality, tooling, firmware, and compliance.
The project needs a simple but explicit operating system:
- Designated project owners on both sides and clear approval gates
- A single source of truth for drawings, specifications, BOMs, firmware, and packaging files
- Written change requests with impact assessments
- A risk register covering technical, supply, schedule, quality, and compliance risks
- Regular milestone reviews ending with clear decisions and next actions
- Certification and labeling requirements defined for each target market
Compliance cannot be added at the end. Batteries, wireless functions, power systems, materials, claims, labels, instructions, packaging, and countries of sale can all change the required tests and documentation. Qualified compliance specialists should confirm the requirements for the actual target markets before mass production.
Who Will Inspect Before Shipment, and How Will Problems Be Closed Out?
“We will inspect before shipment” is not a complete plan. Decide who will inspect, when inspections will happen, which documents will govern them, how units will be sampled, and what happens if an inspection fails.
Inspection may be performed by the brand's staff, an independent third party, or a qualified professional accepted by both sides. What matters is whether the inspection follows standards already approved by the brand instead of relying on a generic checklist.
A complete inspection and release plan should include:
- Inspection checkpoints from EVT and DVT through pilot production and final shipment
- Inspection location, sampling plan, and defect-acceptance criteria
- Functional, cosmetic, packaging, labeling, and quantity checks
- Required evidence, such as photos, serial numbers, or lot information
- The person authorized to approve, hold, or reject shipment
- Corrective-action deadlines and reinspection rules
Finding a problem is only the first step. If an inspection finds defects, establish the affected quantity, root cause, interim containment, long-term corrective action, responsible owner, and verification method. Otherwise, the same problem is likely to return in the next production batch.
The lowest unit price does not necessarily mean the lowest risk. A supplier may reach a low quotation through a higher MOQ, less testing, lower-grade components, limited engineering support, or assumptions that have not been stated clearly.
The right question is not “Which factory is cheapest?” but “Which factory and price are appropriate for our current stage?”
Choose the Factory Only After the Decisions Are Clear
Once these five questions have clear answers, supplier selection becomes more accurate. The team can compare factories against the same product requirements, commercial assumptions, quality standards, project process, and delivery responsibilities.
The right manufacturing partner is not necessarily the factory with the lowest quotation or longest client list. It is the factory whose engineering capability, process management, capacity, supply chain, communication, and commercial model genuinely fit the product.
This sequence cannot eliminate every risk, but it exposes risks earlier, creating an opportunity to manage them before the investment grows.
Frequently Asked Questions
What information should we prepare before contacting a factory in China?
Prepare the product's core requirements, target retail price and cost, projected order volume and acceptable MOQ, development schedule, countries of sale, quality requirements, and permitted scope of customization.
Should we use an existing factory platform or develop a fully custom product?
It depends on differentiation, projected sales, available capital, time to market, and technical risk. An existing platform reduces time and investment but limits differentiation. Full customization provides greater control but requires stronger product definition, validation, engineering management, and financial capacity.
When should quality control begin?
Quality control begins with product definition. Review it at every development and production-control milestone so problems are identified and resolved early.
Is a sourcing agent enough to manage the entire product-development process?
No. Sourcing, product, engineering, quality, compliance, and production require distinct capabilities and accountable owners.