Selecting a contract manufacturer without buying future problems

Start with the manufacturing strategy

Before asking manufacturers for quotations, the company needs to understand what it is actually trying to outsource.

That sounds obvious.

It often is not.

Are you looking for:

build-to-print assembly?

PCB assembly?

box build?

full turnkey manufacture?

supply-chain management?

test development?

new product introduction support?

design for manufacture?

design ownership?

repair and refurbishment?

fulfilment directly to customers?

Different manufacturers are good at different things.

A business needing significant industrialisation support should not select a factory designed primarily for stable, high-volume production.

Likewise, a sophisticated NPI manufacturer may be unnecessarily expensive once the product reaches mature volume.

The manufacturing strategy should therefore begin with the product and the expected journey.

Where is the product now?

What volume will be required in twelve months?

What might volume reach in three years?

How stable is the design?

How complex is the assembly?

What regulatory environment applies?

What geographical markets will be served?

How much manufacturing knowledge exists internally?

The answers define the type of manufacturing partner required.

Do not confuse a famous manufacturer with the right manufacturer

Hardware companies can be attracted to large global contract manufacturers because the names feel reassuring.

Scale matters.

So does fit.

A manufacturer operating factories capable of building millions of products may have little commercial interest in a company planning to manufacture 2,000 units next year.

The account may receive limited engineering attention.

Purchasing leverage may be weak.

Production may be fitted around larger customers.

Small changes may become slow and expensive.

The opposite can also happen.

A small manufacturer may provide excellent attention during early builds but lack the systems, capital, buying power or capacity required when the company scales.

The objective is not to find the biggest manufacturer.

It is to find the manufacturer whose capabilities and commercial interests align with the next stage of the business.

Understand how important you will be to them

This is one of the most overlooked considerations in supplier selection.

Ask:

Where will we sit in this manufacturer's customer portfolio?

If you represent 0.02 percent of the manufacturer's revenue, do not assume you will receive the same attention as one of its strategic accounts.

Conversely, representing an enormous share of a very small supplier's revenue creates a different risk.

Your growth could exceed their financial or operational capacity.

The ideal position is usually somewhere in the middle.

Large enough to matter.

Small enough that the supplier can support substantial growth.

This commercial alignment often matters more than another one or two percentage points of quoted unit cost.

Compare total cost, not quoted price

A supplier can quote a low unit price and still be the most expensive option.

The proper comparison is total landed and operating cost.

Depending on the product, that may include:

• Component cost
• Manufacturing conversion
• Test
• Tooling
• NRE
• Programming
• Packaging
• Freight
• Duties and tariffs
• Storage
• Quality inspection
• Scrap
• Rework
• Warranty
• Engineering support
• Minimum order quantities
• Excess material
• Obsolescence
• Payment terms
• Working capital

Suppose one manufacturer is £10 cheaper per unit but requires twelve weeks more inventory.

If the product contains £2,000 of material, the working-capital difference can quickly exceed the unit-price saving.

Similarly, a factory with poor first-pass yield can destroy apparently attractive manufacturing economics.

The quotation is therefore one input into the decision.

It should not be the decision.

Make sure everybody is quoting the same product

RFQs frequently produce misleading comparisons because suppliers have made different assumptions.

One quotation includes full test.

Another assumes customer-supplied test equipment.

One includes packaging.

Another excludes it.

One uses approved component manufacturers.

Another has substituted parts.

One assumes annual demand.

Another prices each individual order.

The resulting spreadsheet gives the appearance of precision while comparing fundamentally different offers.

A good RFQ should clearly define:

the BOM revision,

drawings,

manufacturing specifications,

expected volumes,

build quantities,

test requirements,

quality requirements,

packaging,

material sourcing responsibility,

incoterms,

delivery location,

warranty expectations,

regulatory requirements,

and expected engineering support.

The manufacturer should also be required to identify every assumption and exclusion.

Silence should not be allowed to become an assumption later.

Separate recurring and non-recurring costs

Manufacturing quotations often mix costs that behave very differently.

Recurring costs scale with production.

Non-recurring costs are generally associated with establishing the production process.

Typical non-recurring costs may include:

tooling,

fixtures,

test equipment,

programming,

production engineering,

NPI activity,

line setup,

and qualification builds.

These need to be visible separately.

Otherwise a manufacturer can make the unit price look attractive while recovering substantial margin elsewhere.

Equally, the cheapest NRE is not necessarily desirable.

Good production fixtures and test equipment may significantly improve throughput and quality for years.

The correct question is whether the investment creates useful manufacturing capability.

Understand who owns the supply chain

There are several common material models.

The customer buys components.

The manufacturer buys components.

Or responsibility is split.

None is inherently correct.

Problems arise when ownership is unclear.

If the contract manufacturer buys material, the company needs to understand:

Who selects the suppliers?

Who approves substitutions?

Who negotiates prices?

Does the customer have visibility of component cost?

Does the manufacturer add material margin?

Who owns price variance?

Who is responsible for shortages?

Who owns excess material?

What happens when forecasts fall?

What happens when the design changes?

Can the manufacturer place non-cancellable orders?

Can it purchase beyond firm demand?

A phrase such as "manufacturer responsible for procurement" is not enough.

Procurement creates financial commitments.

Those commitments need defined authority.

Forecast liability needs to be explicit

Forecasts are necessary because long-lead material often needs to be purchased before firm customer demand exists.

That creates risk.

The important question is who carries it.

A manufacturer might ask for a twelve-month forecast and then place purchase orders against it.

If customer demand falls, the hardware company may discover that it owns months of excess components.

That may be commercially reasonable.

But it should be deliberate.

The agreement should distinguish between:

forecast,

authorised procurement,

firm orders,

non-cancellable commitments,

and excess or obsolete inventory.

It should also define what approval is needed before unusual commitments are placed.

For expensive or long-lead components, the customer may want specific written authorisation.

A forecast should not quietly become an unlimited purchasing mandate.

Look closely at component sourcing

The supply chain is often where the true difference between manufacturers becomes visible.

A good manufacturer should be able to explain:

how components are sourced,

which distributors are approved,

how counterfeit risk is controlled,

how date codes are managed,

how lifecycle status is monitored,

how shortages are escalated,

how alternates are proposed,

and how traceability is maintained.

This becomes particularly important during shortages.

A desperate purchasing team can find almost any semiconductor somewhere.

That does not mean it should go into the product.

Grey-market sourcing can introduce counterfeit, damaged, incorrectly stored or poorly traceable components.

The sourcing policy should be agreed before the company is under shortage pressure.

Evaluate engineering capability, not just factory capability

Early production normally exposes design problems.

That is normal.

What matters is how effectively the manufacturer can help resolve them.

An NPI-capable manufacturer should be able to provide meaningful feedback on:

assembly difficulty,

testability,

tolerances,

component availability,

manufacturing processes,

tooling,

cycle time,

yield,

and cost reduction.

During supplier selection, ask to see examples.

What DFM output does the manufacturer actually produce?

Who performs it?

How experienced are the engineers?

What happens when they identify a design issue?

How quickly can engineering changes be implemented?

A manufacturer that simply follows drawings may be perfectly suitable for a mature product.

It may be a poor choice for a product still moving through industrialisation.

Test capability deserves separate scrutiny

A product can be assembled quickly and still be impossible to manufacture economically because test becomes the bottleneck.

Understand:

what production tests are required,

who develops the test process,

who owns the test software,

who supplies test equipment,

how calibration is controlled,

how test results are stored,

whether results are traceable to serial number,

what happens when a unit fails,

and how long the test takes.

Test capacity should also be modelled.

If one test takes thirty minutes and only one fixture exists, the theoretical maximum throughput is less than 16 units in an eight-hour shift before utilisation and failures are considered.

The assembly line could produce 100 units a day.

The factory would still only be capable of shipping around 16.

This is why manufacturing capacity should be based on bottlenecks rather than headline factory size.

Visit the factory

A supplier presentation is not a factory.

Visit the site where your product will actually be made.

Observe what happens rather than only what is presented.

Look at:

material storage,

ESD control,

production discipline,

work instructions,

tool control,

line organisation,

test areas,

quarantine material,

rework,

non-conforming material,

maintenance,

traceability,

and general housekeeping.

Ask operators how they know which revision to build.

Ask how a defective component is handled.

Ask where obsolete material goes.

Ask how engineering changes reach the line.

Ask how test equipment is controlled.

Processes that work tend to be visible.

Processes that exist only in audit documents are much less useful.

Audit the quality system in practice

ISO certification can be valuable evidence of a management system.

It is not proof that the supplier will manufacture your product well.

The more useful questions are operational.

How does the supplier qualify a new process?

How are first articles approved?

How are defects recorded?

How is root cause analysis performed?

How are corrective actions closed?

What happens when supplier quality deteriorates?

How does the manufacturer prevent recurrence?

How is calibration controlled?

How are operators trained?

How does it control rework?

Ask for examples.

A good quality system should leave evidence.

Understand production capacity properly

Manufacturers often provide impressive capacity numbers.

They may not be relevant to your product.

The useful questions are:

What equipment will our product use?

What is the capacity of that equipment?

What other customers compete for it?

What shift pattern is assumed?

What happens during maintenance?

Where is the bottleneck?

What capacity is available today?

What capacity is committed?

What investment is required for our forecast?

Who pays for that investment?

When must the decision be made?

The difference between theoretical and available capacity can be substantial.

If your production plan depends on significant growth, the manufacturer should be able to demonstrate how that growth will be accommodated.

Think about geography beyond labour rates

Manufacturing location affects much more than conversion cost.

It affects:

lead time,

freight,

duties,

tariffs,

currency exposure,

communication,

engineering access,

customer proximity,

supply-chain resilience,

intellectual property risk,

and working capital.

A labour-saving of £15 per unit can be irrelevant if the finished product costs hundreds of pounds more to transport.

For bulky or high-value systems, regional manufacture may have significant benefits.

For lightweight consumer electronics, a highly concentrated Asian supply chain may remain economically compelling.

The right answer depends on the product.

"China is cheaper" or "we should manufacture locally" are not strategies.

They are assumptions that need testing.

Payment terms matter

Payment terms directly affect the amount of capital required to grow.

Consider two suppliers.

Supplier A requires:

50 percent with order,

50 percent before shipment.

Supplier B offers:

60 days from shipment.

Even at an identical product price, the working-capital difference can be substantial.

Material deposits deserve particular attention.

Manufacturers may reasonably require advance funding for expensive customer-specific material.

But leadership should understand the total cash exposure created by the arrangement.

A manufacturing contract is partly a financing arrangement.

It should be evaluated accordingly.

Clarify tooling ownership

Who owns the tooling?

The answer should be obvious.

It often is not.

If the company pays for tooling, the agreement should normally identify:

the tooling,

the owner,

where it is stored,

how it is maintained,

whether it can be used for other customers,

what happens if the relationship ends,

and how the tooling can be transferred.

The same applies to production fixtures and test equipment.

A company discovering that it cannot move its own product because essential tooling belongs to the outgoing manufacturer has created avoidable dependency.

Protect production data and intellectual property

The manufacturing relationship often requires sharing highly valuable technical information.

The manufacturer may receive:

drawings,

Gerbers,

source code,

test software,

BOMs,

manufacturing processes,

supplier details,

and product know-how.

The commercial framework should clearly address ownership and permitted use.

It should also deal with manufacturing data created during the relationship.

Who owns:

work instructions?

fixtures?

test software?

production improvements?

manufacturing process IP?

documentation created specifically for the product?

The company should be able to understand what it can take with it if the manufacturing relationship ends.

Consider how difficult it would be to leave

This may be the most valuable question in contract manufacturer selection:

What happens if we need to move?

Assume that three years from now:

pricing has become uncompetitive,

quality has deteriorated,

the factory cannot support growth,

the supplier is acquired,

or the commercial relationship simply no longer works.

Can the product be transferred?

Does the company have the complete manufacturing data pack?

Does it own the tooling?

Can test equipment be moved?

Are supplier agreements transferable?

Is there unique manufacturing knowledge held only by the supplier?

How much inventory would become trapped?

Does the contract require transition support?

A good supplier relationship should be built with the expectation that it will succeed.

A good contract should still consider what happens if it does not.

Do not source yourself into dependency accidentally

There can be good reasons to use a single manufacturer.

Splitting small volumes across several factories can create inefficiency, duplicated tooling and inconsistent quality.

But single sourcing should be a conscious risk decision.

If the manufacturer controls:

the supply chain,

tooling,

test equipment,

manufacturing documentation,

product knowledge,

and supplier relationships,

then changing partner may become extremely difficult.

The company has effectively outsourced not only manufacturing but its ability to manufacture.

That may be acceptable.

Management should understand that it has done it.

Pilot builds reveal more than audits

A supplier can perform extremely well during qualification and still struggle with the actual product.

The most useful evidence comes from building it.

Early pilot builds should measure more than whether the units eventually work.

Track:

material availability,

build time,

first-pass yield,

defects,

rework,

engineering interventions,

test time,

documentation issues,

supplier responsiveness,

and causes of delay.

Ask how much support came from your own engineering team.

If engineers needed to stand beside the production line continuously explaining how to assemble the product, the build did not demonstrate manufacturing readiness.

It demonstrated engineering-supported assembly.

That distinction matters.

Define how performance will be managed after selection

Supplier selection is not complete when the contract is signed.

The relationship needs a governance structure.

Depending on the scale, that may include:

weekly operational reviews,

monthly performance reviews,

quarterly business reviews,

quality reviews,

forecast discussions,

cost reduction activity,

and executive escalation.

The important thing is that problems become visible early.

Useful metrics may include:

on-time delivery,

first-pass yield,

defects,

customer escapes,

lead time,

material shortages,

cost variance,

forecast accuracy,

inventory exposure,

and corrective-action closure.

Supplier management should increasingly rely on evidence rather than anecdote.

The cheapest supplier can be extremely expensive

A manufacturing partner influences:

product quality,

customer delivery,

cash,

inventory,

engineering workload,

margin,

reputation,

and the company's ability to scale.

That makes the selection decision substantially more important than the difference between two quoted prices.

A good manufacturing partner provides capability that would be difficult or uneconomic for the business to build itself.

A poor one creates an operating constraint that can take years to remove.

Select on price and you may get a cheap product.

Select on total cost, capability, control, scalability and commercial alignment and you have a much better chance of building a manufacturing system that will still work when the business is ten times larger.

The objective is not to find someone who can build the next hundred units.

It is to choose a partner that does not become the reason you cannot build the next hundred thousand.