Comparison · make or buy · in-house production and fulfilment versus outsourcing · equipment, staffing, inspection, fixed versus variable cost
This is the decision that ends careers as well as budgets. Making merchandise in-house means buying machines, finding a room with the power and the ventilation to run them, hiring operators who already know how to set up a press, and accepting that the person who loaded the machine is also the person who inspects the result. Outsourcing it means giving up all of that and paying per unit for a process somebody else has already refined. The comparison is usually framed around cost per item, which is the least useful number available, because a workshop that sits idle for eleven months is not cheap at any unit price. What actually decides it is volume, regularity, method range and the presence or absence of a deadline. An organisation printing two thousand identical screen-printed tees a year, from one design, through one method, is in the best possible position to make them itself. An organisation needing four decoration methods, four deadlines and a hundred different specifications is paying to buy access to capability it will never use. This page covers what in-house production really involves, what staffing and inspection have to look like, where the fixed cost sits, and the specific volume and regularity thresholds at which each side of this decision is correct.
The phrase covers a wide range of commitments, and buyers routinely underestimate the lower end of it. Producing apparel decoration in-house means a screen printing table or carousel with a flash dryer, or a garment printer with a pre-treatment unit and a curing press, or an embroidery machine with a stabiliser frame. Each has a footprint, a power requirement and a ventilation problem. Screen printing needs space for ink mixing, screen storage and drying; ink and solvents are a workplace hazard that requires handling procedures, storage and disposal planning. Garment printing needs power for the dryer and heat for the press, and a floor flat enough for the tunnel. Embroidery needs thread inventory, which is bulky, and an operator who can set a digitising file and read a machine error.
Then there is the machine that nobody planned for, which is the finishing and packing capability. A heat press for transfers and cut-and-sew items, a sewing machine for simple assembly, an embroidery finishing station for garment preparation, a laser or engraver for acrylic and metal, and a packing table with the consumables to ship safely. Very few organisations that start producing in-house end up with a single method, because the first order raises a question the current equipment cannot answer, and the answer costs money.
Alongside the hardware there is a maintenance obligation that does not pause. Screens degrade, print heads clog, bearings wear, heaters fail and software licences expire. Consumables have to be reordered before the last unit runs out, which means somebody has to know the reorder points. And the equipment is capital: it depreciates, it needs insurance, and it occupies floor space that could have held desks or stock. None of that appears in a unit-cost comparison, and all of it is real.
A dated, small-volume requirement that is usually needed in a hurry and is judged by people looking at it from two metres away. This is the worst candidate for in-house production at low volume, because setting up a press for forty personalised garments costs more in setup time than the garments cost to make elsewhere. It is a reasonable candidate at several hundred units of one design, where the same screen and the same setup absorb the cost across the run.
The one category where in-house production genuinely pays, provided the design is stable. Thousands of identical units through one screen or one press is the workload that amortises a machine over a year. The requirements are narrow and the discipline is total: a frozen specification, a scheduled maintenance regime and a quality check that is not done by the person loading the machine.
Small quantities across a wide range of items, which is the hardest pattern for in-house production. Each item type is a different production method, each run is short, and the setup time dominates the run time. This category needs the range of a supplier: they already own the equipment for the item you did not think of, and the setup for a fifty-unit run is already amortised across other work.
Small, personalised, often embroidered, and assembled by hand to a fixed pack-out. Embroidery in particular is a skilled trade with a real learning curve, and the failure mode of a self-run first attempt is legible to the recipient: uneven stitch density, a dropped thread, a name spelled from a list that was not reconciled. Buying this is the straightforward decision.
Equipment can be specified against a requirement. Staffing cannot, because the requirement is a set of judgement calls rather than a task list. A screen printing operator does not merely load garments onto a press. They ink the colours, mix to match a standard, register the screens, judge the flash time, watch for a registration drift developing across a run and stop the run when it drifts. That judgement comes from seeing it fail, and it takes months to build.
The arithmetic is uncomfortable. Suppose a method needs one operator and the organisation runs apparel orders intermittently, perhaps six weeks a year. For the rest of the year that operator is a cost with no output. The alternative is to keep the person on other work and contract the six weeks, which is exactly what an outsourcer does. Where the work is steady, the in-house model is efficient, because the operator's judgement improves with the volume and the setup time per run falls as the process becomes familiar.
Small organisations hit a second problem, which is that one person cannot be two people. The person who prints cannot also do the pre-treatment, the finishing, the folding, the kitting and the despatch, and in a small team somebody ends up doing all of it in the wrong order. The result is a queue at the packing table while the press is idle, or a garment that is folded before it is cured. In-house production works when the stages have separate owners, and it degrades quietly when they do not.
Recruitment is the third constraint. Printers, embroidery operators and laser operators are a small labour market, and an organisation competing with established producers for the same people is competing from a weak position on pay and from no position at all on throughput. This is the case for outsourcing that is strongest when the volume is lowest, which is the opposite of where the capital-cost argument points.
An in-house quality check is performed by the person who made the item, usually within minutes of making it, usually knowing exactly which run went wrong and mentally excusing it. That is not dishonesty, it is the ordinary human response to finishing something after a long day. The inspection still has value, but it is not independent, and its value falls with familiarity: the operator who sees the same registration drift every week will stop mentioning it, because mentioning it every week is more annoying than the drift.
An external producer runs inspection as a role, with a sampling plan, a defined reject rate and a record. Before despatch there is normally a final inspection on a sample drawn from packed cartons rather than from the head of the run, which is the version that reflects what the recipient receives. Where the volume is large enough to matter, buyers can and should ask what that plan is, what the agreed reject rate is, and what happens to a consignment that fails it. Those are the three questions that separate a production process from a printing session.
The failure modes differ too, which matters when you are deciding whether to bring work inside. In-house, the escape usually happens at the end: an order is packed with pieces from a bad part of the run, or a whole run is slightly off and nobody measured it, or the count is short because the last pallet was never counted. Outsourced, the escape is more likely to be a quantity or specification mismatch between the purchase order and what the plant thought it was making, which is a communication failure rather than a manufacturing one.
The most useful quality control measure does not depend on either model: hold the specification, measure against it, and check a sample from the packed cartons rather than the finished pile. That works from outside as well as inside. It also forces the same discipline in both, which is worth doing before deciding where the work sits.
The economic argument is a straight one and it is easy to get wrong by comparing the wrong things. In-house production converts a fixed commitment into a set of costs that do not move with volume: the machine, the room, the operator salaries, the maintenance contracts, the insurance. Outsourcing converts the same requirement into a variable cost that scales with the units ordered. Below a certain annual volume the variable model is cheaper, because the fixed commitment is spread over few units. Above it, the fixed model becomes cheaper, because the marginal cost of the hundredth unit is close to the cost of the first.
The threshold is not a universal number, because it depends on the method. Cheap, manual processes cross over at lower volume than expensive automated ones. A screen print table is a modest commitment; a multi-head embroidery machine and a garment printing tunnel with a dryer are not. A buyer working through the arithmetic needs to include the space, the power upgrade, the extraction, the training and the supervisory time, because omitting the supervisory time is the most common error and it is the largest hidden line.
What the arithmetic hides is the capacity trap. In-house capacity is all-or-nothing in practice. In the month before an event the equipment is the bottleneck and three people are idle in October. Outsourcing absorbs that unevenness without a surcharge, because the supplier's capacity is spread across many clients, which is the actual service being purchased when you buy outside: the ability to be busy on demand, with the printing attached.
The trap also runs the other way. At sustained high volume, an in-house operation eventually cannot keep up, and the consequence is either a missed date or a rushed run with worse quality. Businesses that make merchandise internally frequently reach this point and respond by buying equipment again, adding a second method, at which point the fixed cost has doubled and the trap has widened. Recognising that point early is more useful than calculating the crossover precisely.
| Cost or risk | In-house production | Outsourced production |
|---|---|---|
| Cost shape | Fixed: paid whether or not there is work | Variable: scales with the units ordered |
| Best volume | Steady and high, one design repeated | Low, irregular or spiky |
| Method range | One or two methods well; more needs more machines | Any method, already installed |
| Capacity on a deadline | Fixed; a rush is physically impossible | Elastic, with advance booking still required |
| Operator skill | Months to build, and hard to hire | Already present |
| Inspection independence | Low; the maker inspects | Higher; inspection is a separate role |
| Speed of a change | Fast if an operator is free | Fast until the slot is booked |
| Fixed commitments | Equipment, space, power, ventilation, maintenance, supervision | None, beyond the relationship |
| Main risk | Underutilised equipment and an unavailable operator | Supplier capacity and lead time |
Six situations make internal production the wrong answer regardless of how the arithmetic looks, and four of them are about capability rather than cost.
Low or irregular volume. Six weeks of work a year does not support a workshop. The fixed cost is being paid to produce nothing for most of the year, and the operator's judgement never accumulates because there is no repetition to accumulate from.
More than one or two decoration methods. Every additional method is another machine, another operator, another consumable chain and another set of maintenance obligations. A supplier has all of them installed and amortised, and this is usually the single strongest argument for buying outside.
Hard external deadlines with a spiky calendar. Events cluster. In-house capacity cannot absorb a cluster, and the rush that follows is where quality escapes. Book capacity outside instead, early, before the cluster arrives.
No local operator talent. Where the labour market is thin, recruiting is slow and the cost is high, and where the existing staff already have full responsibilities. This is common in organisations whose core business has nothing to do with manufacturing.
Certification and compliance requirements. Restricted-substance controls, documented testing and audited processes are expensive to build and trivial to buy from someone who holds them. Rebuilding them internally to satisfy a procurement questionnaire is a two-year project that nobody budgets for.
Capital and space are constrained. Floor space in a headquarters is expensive and a workshop needs power, extraction and a fire-safe storage arrangement. The opportunity cost of the room alone often exceeds the whole outsourcing cost for several years.
The common thread is that in-house production is a bet on continuity. Steady, repeated, narrow work justifies it. Everything else is a bet that does not pay, and the machine sitting idle is the visible cost of having taken it.
Four situations favour making it internally, and they are fewer than most organisations assume.
High volume of one design through one method. This is the case the model exists for. Thousands of identical units amortise the equipment, the setup and the training, and the operator's speed improves until the setup time is negligible. Nothing beats this structure, and an outside supplier has to earn its margin on exactly the same arithmetic.
Hard internal confidentiality requirements. Where the artwork cannot leave the building for contractual or regulatory reasons, internal production may be the only option. Check whether the requirement is really that restrictive: many organisations assume it and find that a supplier under a written confidentiality undertaking satisfies the requirement entirely.
Very short iteration loops on a small but continuous run. If the design changes weekly across a continuing programme, physical iteration in-house is faster than the round trip to a supplier, and the print quality can be judged by touching it rather than by describing it.
Existing capability already in place. If the organisation has a workshop, an operator and a maintenance regime, the marginal cost of doing the next order there is far lower than starting a supplier relationship. The decision is then about utilisation, not about capability.
The honest framing is that in-house production is a strategic choice about how the organisation spends its capital and its management attention. It is a good choice for a narrow, steady, high-volume workload and a poor one for anything else. Most organisations that over-invest here have done it because the first small order was urgent and the machine was on the way.
The verdicts assume a B2B buyer in Armenia or the wider EAEU with an irregular, event-driven merchandise programme. Where the answer turns on a fact the buyer holds, that fact is named.
Outsource until one design repeats at high volume. Do not buy equipment on the strength of a single order, however large. The repeat is what amortises the machine, and the second occurrence of the same design is the earliest sensible point to reconsider.
Count the supervisory time before deciding. Add the hours per week that somebody spends managing the workshop, ordering consumables, servicing machines and chasing deliveries. Excluding it is the most common error in the arithmetic, and it is usually the largest line.
Outsource anything involving a method you do not already own. A second machine is a second commitment with a second operator and a second maintenance regime. Rent the capability first, and buy it when the volume justifies the whole stack rather than one part of it.
Book external capacity before the season, not in it. The deadline that matters is not the production date, it is the date the slot has to be reserved. Suppliers plan backwards from that, and the organisations that lose their dates are the ones that order in the same week as the event.
Insist on a pre-despatch sample inspection regardless of the model. Drawn from packed cartons, against the written specification, with the reject rate agreed in advance. It is the only quality measure that works identically whether the work is inside or outside.
Do not mix the two for the same design. Running one specification through an internal press and an external supplier in the same month produces two different results and a permanent argument about which is correct. Pick a source per design and hold it.
There is no universal number because the threshold depends on the method, and manual processes cross over at a lower volume than expensive automated ones. The way to find your own figure is to take the full fixed cost of the capability: equipment, space, power and extraction, operator salaries, maintenance contracts, consumables overhead and the supervisory time, then divide it by the variable cost of an equivalent external order. Above that volume the internal route is cheaper, below it the external route is. Most garment decoration crosses over somewhere in the low thousands of units a year for a single design, which is why the same organisation can sensibly make one programme internally and buy the rest outside.
Treat it as a skilled trade rather than a production role. The work is judgement under time pressure: matching a colour, watching for registration drift, reading a machine fault, deciding when to stop a run. People with that judgement are scarce and they are usually employed by producers who can offer steadier work and faster machines. If the volume is intermittent, there is no honest way to compete for them, and the correct decision is to contract the skill rather than to hire it. If the volume is steady, retention comes from operating decent equipment, keeping the working day predictable and paying for the training that makes the operator faster.
Three things, in this order. A written specification the factory works from, with garment, colour reference, decoration method and position dimensions written down. A stated pre-despatch inspection plan, normally a sample drawn from packed cartons rather than from the head of the run, with an agreed reject rate. And a record that travels with the consignment, so a problem can be diagnosed rather than argued about. Also agree in advance what happens if the sample fails: replacement, rework or credit, and on what timescale. A supplier that answers all four questions concretely is telling you something real about how the work is run.
No. Corp-Merch is an independent B2B merchandise supplier. References to equipment categories, certification schemes, standards organisations and trade bodies named on this page are informational only and describe the general shape of the category. They imply no sponsorship, endorsement, affiliation or official-supplier relationship with any of them, and no trademark is claimed as owned by Corp-Merch.
This page is written for B2B buyers of corporate merchandise in Armenia and the wider EAEU weighing internal production against outsourced production and fulfilment. Corp-Merch is an independent B2B merchandise supplier. Statements about equipment, staffing, inspection practice and cost structure describe how this work is typically organised in the category and are not a claim about any named manufacturer, plant or standards body. Every order is quoted individually against a written specification, and nothing on this page is a binding price, a capacity guarantee or a commitment to any lead time.