The choice between running an OEM programme and operating your own workshop is a choice about where capital and risk sit. An OEM programme converts fixed cost into variable cost: you buy capacity, tooling and expertise as needed, at a standard MOQ of 500 pieces per style per colourway, with sampling in 6–10 working days and bulk in 35–50 days. In-house production gives control over sequencing, iteration and intellectual property but demands premises, equipment, skilled staff and volume to absorb them. Most scaling brands use OEM first and add selective in-house capability later.
Two Operating Models, Two Cost Structures
The distinction is not really about geography or about who owns a building. It is about which costs are fixed and which move with volume. Understanding that difference makes the decision straightforward, because the right answer depends on how confident you are about future volume.
| Dimension | OEM programme | In-house DTC production |
|---|---|---|
| Capital profile | Mostly variable — buy what you need | Mostly fixed — premises, equipment, payroll |
| Unit economics at low volume | Better; no idle capacity to absorb | Worse; fixed cost spread over few units |
| Unit economics at high volume | Margins shared with the partner | Better once fixed cost is absorbed |
| Speed to first product | Fast; capability already exists | Slow; capability must be built |
| Iteration speed | Sample cycle per round | Immediate; changes tested same day |
| Control over IP and process | Depends on contract terms | Direct and structural |
| Scalability ceiling | High, subject to partner capacity | Limited by your own recruitment and space |
Programme parameters at a glance
| Parameter | Commitment |
|---|---|
| Minimum order quantity | 500 pieces per style / colourway |
| Sampling lead time | 6–10 working days (12–15 days for complex constructions) |
| Bulk production lead time | 35–50 days after sample approval |
| Inspection standard | AQL 2.5 (ISO 2859-1), single sampling, general level II |
| Incoterms | FOB Xiamen — EXW, CIF and DDP available on request |
| Payment terms | T/T 30% deposit / 70% before shipment; L/C at sight for container loads |
| Verified production floor space | 4,950 m² (SGS-verified) |
| Team | 137 staff across 7 production lines |
| Equipment | 149 machines including computerised saddle-stitch and edge-painting units |
| Monthly capacity | 200,000 pieces |
| Entity | Founded 2014; principal has worked in bag development since 2004 |
Note the crossover. In-house production becomes economically attractive only above the volume where fixed costs are absorbed, and that threshold is higher than most first-time estimates assume. Below it, the same bag built by a production partner is cheaper, faster and usually better made, because the partner's operators build that construction every week rather than intermittently.
It is also worth separating control from ownership. A brand working through a programme retains control through the documents it owns — the tech pack, the bill of materials, the tolerance table, the defect dictionary and the retained approved sample. What it delegates is execution. Confusing the two leads brands to build workshops for reasons that a better brief would have addressed.
What Each Model Demands From You
Neither model is passive. They demand different capabilities from the brand team, and choosing against your strengths is a common cause of a stalled scale-up.
| Capability | OEM programme | In-house production |
|---|---|---|
| Technical specification | Essential — must brief precisely | Essential — and must draft patterns |
| Pattern making | Useful for review, not required | Essential core skill |
| Production planning | Coordination and checkpoint management | Line balancing, scheduling, output control |
| Quality systems | Define acceptance criteria and review reports | Build and run the whole system |
| Equipment maintenance | Not required | Ongoing operational responsibility |
| Skilled labour recruitment | Not required | Critical and often the binding constraint |
| Compliance and audit | Verify partner evidence | Hold and maintain your own |
- If your strength is design and brand, an OEM programme lets you stay there without building a manufacturing organisation.
- If your strength is craft and process, in-house production turns that into a defensible asset rather than a service you buy.
- If you have neither yet, OEM is the lower-risk route — mistakes cost a sampling cycle rather than a payroll.
- If intellectual property protection is the dominant concern, in-house removes the counterparty risk entirely.
Capital: Where the Money Actually Goes
In-house production is often presented as a way to capture the factory's margin. That is true in theory and frequently untrue in practice, because the margin is consumed by fixed costs that only disappear when volume is sufficient to absorb them.
- Premises: rent, utilities, fit-out, insurance and the cost of space that cannot be repurposed quickly.
- Equipment: purchase, maintenance, calibration and eventual replacement of machines that sit idle between orders.
- People: recruitment, wages, training and the disruption of turnover among skilled operators.
- Inventory: material held continuously, including slow-moving colours and hardware that may never be used.
- Compliance: your own audits, certificates and testing programme rather than verification of a partner's evidence.
| Cost line | OEM programme | In-house production |
|---|---|---|
| Premises | None | Rent, utilities, fit-out, insurance |
| Equipment | Partner-owned; tooling charged per style | Purchase, maintenance, calibration, replacement |
| Skilled staff | Partner-employed | Recruitment, wages, training, turnover risk |
| Material inventory | Procured per order | Held continuously, including slow-moving stock |
| Compliance and audit | Verify partner records | Own audits, certificates and testing programme |
| Idle capacity | Absorbed by the partner | Borne directly by you between orders |
The idle capacity line is the one most often missed. A partner's line moves from programme to programme; yours waits for your next order. Unless your volume is consistent enough to keep skilled operators productively occupied, in-house capacity becomes an overhead rather than an asset.
Speed, Flexibility and Iteration
In-house production wins decisively on iteration. A change that requires a sampling cycle externally can be tested on the bench the same afternoon internally. For a brand still finding its design language, that speed has genuine value — provided someone on the team can execute it well.
- Define what you are iterating on. Design details benefit from in-house speed; material and hardware options do not.
- Measure iteration value against operator skill. An in-house change executed badly teaches you less than an external sample executed well.
- Account for sampling cycle time honestly. Six to ten working days per round externally is real, but so is the time your own team spends.
- Consider a split: development samples in-house, bulk production through a programme partner.
- Revisit the balance each season, since the right answer moves with your volume.
Where speed matters most is at the extremes: very early, when the design is unformed, and very late, when a reorder is needed quickly. In the middle, when the specification is stable, the iteration advantage largely disappears, and the two models converge on the same limiting factor — how quickly a line can be set up and run.
Where iteration speed genuinely pays
- Proportion and volume: small changes read very differently in three dimensions than on a screen.
- Handle and strap geometry: comfort is difficult to specify and easy to assess on a physical piece.
- Internal organisation: pocket depth and placement are best tested with the items they must hold.
- Closure and hardware interaction: how a lock engages under load is rarely resolved on paper.
- Construction sequence: where a seam becomes hard to execute is often discovered only by building it.
Quality Control Under Each Model
Quality control is often assumed to favour in-house production, on the logic that you can watch it being made. In practice, watching is not the same as controlling, and a partner with a documented inspection system frequently outperforms an owner-operator with good intentions and no records.
The requirement that travels across both models, and the one most often missing, is the written standard. Whether inspection is performed by a partner's team or your own, an inspector without a defect dictionary and a retained sample is making decisions you cannot audit.
The Hybrid Path Most Scaling Brands Follow
In practice the decision is rarely binary. Brands typically begin with an OEM or ODM programme, prove demand, then add narrow in-house capability for the work where iteration speed or confidentiality matters most — often prototyping, customisation or repair.
| Stage | Typical arrangement | Rationale |
|---|---|---|
| Launch | ODM or OEM programme | Capability exists; capital stays in inventory and brand |
| Validation | Small-batch pilots through the programme | Tests demand without fixed-cost commitment |
| Growth | OEM programme with in-house prototyping | Fast iteration on design; no fixed production overhead |
| Scale | OEM programme plus in-house customisation or aftercare | Captures the highest-touch work; bulk stays variable |
| Maturity | Balanced split reviewed annually | Optimises unit economics against capital and flexibility |
What changes as volume grows
The economics of both models shift with scale, and the shift is rarely smooth or predictable at the outset. Understanding which cost lines behave non-linearly prevents the most common planning error: assuming that the unit cost achieved at one volume will hold at the next.
| Cost line | Behaviour as volume rises | Planning implication |
|---|---|---|
| Material unit cost | Improves with vendor volume tiers, then flattens | Model price breaks at realistic thresholds rather than assuming continuous improvement |
| Tooling | Amortises across more units; fixed per style | Custom tooling becomes reasonable once a style has reordered |
| Labour minutes per unit | Fall with repetition, then plateau | Do not assume the learning curve continues indefinitely |
| Fixed overhead | Absorbed per unit as volume rises | The main argument for in-house, but only with consistent demand |
| Inspection cost | Roughly proportional to volume | Does not improve with scale; budget it as a per-unit cost |
| Freight | Improves per unit with consolidation | Mode and consolidation matter more than volume alone |
Two of these lines deserve particular attention because they are routinely mispriced. Labour minutes per unit fall sharply in the first few hundred pieces as operators learn a construction, then plateau — programmes that extrapolate the early curve into later orders consistently under-quote. Inspection cost, by contrast, does not improve with scale at all; it is a per-unit cost that must be budgeted as such rather than treated as an overhead that dilutes.
A Framework for Deciding
Five questions usually settle it. Answer them honestly against your current position rather than your projected one, and revisit the list annually, because the answers shift as volume, team capability and capital position change over successive seasons.
- Is your volume consistent enough to keep skilled operators productively occupied year-round?
- Does your team include pattern-making and production-planning capability, or would you be hiring it?
- Is intellectual property protection a dominant concern that contracts alone cannot address?
- Do you need to iterate on construction several times per week, or is your specification stable?
- Can you carry the capital cost of premises and equipment without constraining inventory or marketing?
If most answers point to variable cost, an OEM programme is the more rational route. If they point to control and consistent volume, selective vertical integration is worth modelling. Neither answer is permanent, and brands that revisit the question each season tend to make better decisions than those that treat it as a one-time strategic commitment.
Risk: What Each Model Exposes You To
Every operating model concentrates risk somewhere. The purpose of naming it is not to avoid it — the purpose is to know which risk you are carrying, so you can manage it rather than discover it.
| Risk | Under an OEM programme | Under in-house production |
|---|---|---|
| Design leakage | Managed by contract and access control | Structurally low; remains internal |
| Quality escape | Managed by inspection scope and defect definitions | Depends entirely on your own systems and discipline |
| Capacity shortfall | Partner capacity, subject to forward booking | Your recruitment, space and equipment |
| Cost volatility | Shared; sensitive to material and currency movement | Borne directly, including idle capacity |
| Key person dependency | Lower; skills sit with the partner | Higher; skills sit with named individuals |
| Compliance exposure | Verify partner evidence and test reports | You hold the certificates and testing programme |
| Capital lock-in | Low; tooling is the main committed cost | High; premises and equipment are committed |
Notice that the two columns are close to inverse. The OEM programme concentrates risk in the counterparty and in shared economics; in-house production concentrates it in capital and in people. Which is preferable depends on which you can manage better, not on which sounds safer.
Optionality is the third consideration. A programme can be changed, scaled or moved with months of notice. A workshop cannot. Where demand is uncertain, the option to adjust has value that a unit cost comparison does not capture, and brands that ignore it tend to discover its worth at the worst possible moment.
Model your scaling options with the programme team
Share your current volume, target growth and capital constraints, and we will help you compare an OEM programme against selective in-house capability. Email service@junyuanbags.com or WhatsApp +86 177 5002 0688.
Frequently Asked Questions
What is the difference between OEM and DTC manufacturing?
In an OEM programme you direct production through a partner who builds to your specification under your brand, converting most cost into variable cost. In DTC or in-house manufacturing you own the premises, equipment and payroll. The models differ mainly in where fixed cost, capital and control sit rather than in the physical work.
Is in-house manufacturing cheaper than using an OEM programme?
Not at low volume. In-house production adds premises, equipment, maintenance, recruitment and idle capacity, all of which must be absorbed before unit cost improves. The crossover point depends on consistent volume, construction complexity and local labour cost. Below that volume, a programme partner is usually cheaper and often better made.
When does it make sense to bring production in-house?
When volume is consistent enough to keep skilled staff productively occupied, when iteration speed materially affects the product, or when intellectual property protection is a dominant concern. Many brands start with prototyping or customisation rather than full production, which captures most of the benefit without the overhead.
Can I use both models at the same time?
Yes, and most scaling brands do. A common arrangement is in-house prototyping and sampling with bulk production running through a programme partner. Another is in-house customisation or aftercare with standard styles produced externally. The split should be reviewed as volume and product complexity change.
How does intellectual property protection differ between the models?
In-house production removes counterparty exposure structurally, since no third party holds your patterns or tooling. A programme model relies on contract: a properly drafted NNN agreement, written ownership of tooling and pattern files, and operational controls on file distribution and branded rejects.
What hidden costs come with in-house bag production?
Idle capacity between orders, equipment maintenance and calibration, recruitment and turnover among skilled operators, and material inventory held continuously including slow-moving stock. Add the compliance and testing programme you must hold yourself rather than verifying from a partner, plus the management time that production absorbs from brand work.
How quickly can an OEM programme scale compared with in-house?
Considerably faster at the top end. A programme partner with multiple lines and established material vendors can absorb a volume jump, whereas in-house scaling requires recruitment, training, equipment and space. That responsiveness is a real commercial asset during a strong season, and it is invisible in a unit price comparison.
Does in-house production give better quality control?
Not automatically. Control comes from a defined standard — a defect dictionary, reference boards, retained samples and a corrective action log — not from proximity. A partner running documented inline inspection with dated closure records will typically outperform an owner-operator relying on observation without records.
What is the minimum volume for in-house production to make sense?
There is no universal figure. The test is whether consistent demand keeps skilled operators productively occupied through the year, and whether the fixed costs are absorbed at your margin structure. Model it against the standard 500-piece MOQ of a programme before committing to premises or equipment.
How does sampling differ between the models?
Through a programme, sampling runs 6–10 working days, or 12–15 for complex constructions, and covers pattern development, material preparation, assembly and finishing. In-house, the same work competes with production for bench time, though iteration between rounds is faster once a pattern exists.
Should a startup brand set up its own workshop?
Rarely. Capital is usually better deployed in inventory, brand and customer acquisition, while a programme provides proven capability and capacity. Building a workshop before demand is proven converts a flexible cost into a fixed one at precisely the moment flexibility matters most.
What capabilities should I keep in-house even with an OEM programme?
Specification writing, material and hardware selection, sample review and quality standard definition. These are the decisions that determine the outcome, and they are difficult to delegate without losing control of the result. Prototyping is worth adding once the volume justifies the bench time and the skills exist on the team.
How often should I revisit the make-or-direct decision?
Annually, or whenever volume, product complexity or capital position changes materially. The right answer moves as the business moves, and a decision that was correct at launch rarely stays correct indefinitely. Brands that treat it as a standing question rather than a one-time commitment consistently make better-timed choices.
Can a programme partner support customisation at low volume?
Often yes, particularly monogramming, colourway extensions and hardware substitutions on an existing base. Small-batch pilots are available below the standard 500-piece minimum, though unit cost is higher at that level. Discuss the specific requirement before assuming it fits a standard production flow, since some customisation needs dedicated tooling.
What should I provide to compare the two routes fairly?
Current annual volume, forecast growth, product mix, capital available for equipment and inventory, target unit cost, and the skills already present in your team. With those, both routes can be modelled on the same basis rather than compared on headline unit price alone.
Does vertical integration improve speed to market?
For iteration, yes. For first production, usually not, since capability must be built before it can be used. The advantage appears once patterns, tooling and skills exist and the constraint becomes bench time rather than communication cycle time. Before that point, a programme partner will typically reach bulk faster.
How much capital does setting up a small bag workshop require?
It varies widely by location, equipment specification and construction type. The lines that most often surprise new owners are not the machines but the ongoing ones: premises, skilled wages during slow periods, maintenance and calibration, and material inventory held continuously. Model the fixed costs over a full year rather than against a single production run.
What volume justifies bringing production in-house?
The test is consistency rather than peak. If demand keeps skilled operators productively occupied across the year, fixed costs are absorbed and unit economics improve. If volume arrives in bursts with quiet periods between, idle capacity usually outweighs the margin captured. Compare against the standard 500-piece MOQ before committing.
Can I start in-house with just finishing and assembly?
Yes, and it is a sensible intermediate step. Many brands keep cutting and heavy preparation with a production partner while doing final assembly, edge finishing or customisation in-house. This captures iteration speed and confidentiality on the most visible operations without committing to the full equipment and staffing load.
Does using an OEM programme mean I lose control of my product?
Not if the programme is specified properly. Control sits in the documents you own: the tech pack, the bill of materials, the tolerance table, the defect dictionary and the retained approved sample. What you delegate is execution, not definition. The brands that struggle are those that delegate both.