One factory that does everything well is the most comfortable arrangement in sourcing, and the most fragile. It holds right up to the moment it does not — the season the line is full, the audit that fails, the reorder that waits behind a client many times your size. Concentration risk stays invisible while things are calm, then presents its bill in your busiest month. This guide is about what betting a brand on a single factory actually exposes, why a captive order quietly loses its leverage, and what a coordinated network of specialists changes about it — including the real costs of the alternative and the cases where one factory is still the right answer.
The short answer
- Single-factory dependence is a continuity risk before it is a quality one. A factory can make a flawless sample and still be full, newly expensive, or unavailable at the exact moment you need it — and none of that shows up in the sample you approved.
- A captive order has almost no leverage. Once a factory knows you have nowhere else qualified to go, price, priority and lead time drift in its favour — not out of malice, but because that is simply how scarce capacity gets allocated.
- A managed network spreads the exposure. A coordinating partner can requalify and reroute your order to another qualified specialist against the same sealed standard, and keep a fabric alternative approved in parallel, so one failure does not stop the program.
- Rerouting is neither instant nor free, and one factory is genuinely right for some brands — a single simple product, a deep and proven relationship, or volumes large enough to make you a priority client. The aim is to choose the dependence, not drift into it.
The specific ways one factory fails you
Start by noticing that almost none of these failures is about quality. A factory that makes a beautiful sample can still deliver none of it this quarter. The continuity failures cluster into a short list: the line is full through peak and cannot take your reorder; the price rises once your patterns, tooling and approvals all live in one building; a compliance audit at that facility fails and every style you make there is suddenly in question; your order is quietly deprioritised behind a much larger client; or the plant closes for a holiday stretch, a relocation, or a machine that went down at the wrong time.
Two of these do the most damage to small and growing brands. Capacity is the first: a factory gives its best weeks to its largest customers, and a brand ordering near a minimum is the order that gets moved when something bigger lands. Deprioritisation is the same force in slower motion — you are not refused, you are rescheduled, and then rescheduled again, while the factory protects the accounts that fill more of its calendar. Neither arrives as a clear “no”. They arrive as a date that keeps sliding.
The other thread is timing. None of these risks is visible in the sample, the price list, or the first smooth order. They surface only after you are committed — after the season’s approved lab dips, the graded patterns and the sealed references all sit in one place. That is the quiet trap of single-factory sourcing: the arrangement feels safest at exactly the point where it is becoming your single point of failure.
Why a captive order has no leverage
A captive order is one with nowhere else to go, and both sides know it. The moment a factory understands that your patterns, your approvals and your only sealed reference live with them and nowhere else, the negotiation has quietly ended. You can still ask for a keener price, a firmer date or a smaller minimum — but you are asking, not negotiating, because you have given up your own ability to walk.
This rarely looks like a villain. It looks like a price that creeps up at each reorder, a quote that takes a little longer to come back, a little less willingness to flex on a minimum or split a run. A factory with finite capacity allocates it rationally, and the customer who cannot leave is the safe one to keep waiting. A brand ordering near a stock minimum of a hundred units is not going to out-argue a client filling the line for a month — not because anyone is being punished, but because that is how scarce weeks get assigned.
The buyer-side reading is that leverage was never about threats or volume. It is about a real, qualified alternative existing at all. You do not have to move your order to benefit from being able to — a factory quotes and schedules differently for a customer it knows could requalify elsewhere. That optionality is precisely what a captive arrangement destroys, and precisely what a network exists to preserve.
Peak season is when concentration bites
Concentration risk is seasonal. For most activewear brands demand compresses into the same narrow windows — the pre-season build, and the post-launch reorder when a style unexpectedly sells. Those are exactly the weeks every other brand at your factory also wants, which means the capacity you need most is scarcest at the very moment you reach for it.
A single factory cannot solve this for you, because it faces the same compression from every customer at once. Your reorder — the one chasing a style that is working — becomes the hardest order to place, and a slipped date here is not a logistics inconvenience but weeks of a selling window you do not get back. The style is selling now, and “we can start in a few weeks” is a very different business answer than it would be in a quiet month.
A network changes the shape of the problem, because the same construction is qualified at more than one specialist. Capacity becomes a pool rather than a single line: when one partner’s calendar is full, a coordinating partner can place the run with another already qualified for that construction and held to the same standard. It does not conjure capacity out of nothing, and it is not instant — but it turns “wait for the one factory” into “route to the next qualified one”, which in peak season is often the whole difference.
How a managed network spreads the risk
A managed network is not “more factories” for its own sake. It is a deliberate arrangement in which the constructions you depend on have more than one qualified home, and the party coordinating them holds the standard rather than any single producer. The working unit is not the factory; it is the qualified route — this construction, at this specialist, against this sealed reference, with at least one more specialist requalified to take it if needed.
Fabric deserves the same treatment, because a mill is its own point of failure. A dye lot that cannot be matched, a base fabric quietly discontinued, a mill that is itself full — any of these can halt a program even when the garment factory is ready. Keeping a second fabric approved in parallel — tested and signed off before you need it — turns a material problem into a substitution rather than a standstill. It is unglamorous work that only ever pays off on the day something breaks.
This is where continuity meets the sealed standard, without being the same discipline. Holding one specification across factories — the reference sample, the tolerances, who inspects and against what — is its own subject. What matters for continuity is where that standard lives. If the only working definition of “correct” is the muscle memory inside one factory, you cannot reroute without starting over. If the coordinating partner holds the sealed reference and the specification, a second qualified specialist can be held to the same bar — which is what makes rerouting a real option rather than a hopeful one.
What rerouting actually costs
None of this is free, and a buyer-side guide should not pretend otherwise. A second factory qualified for a construction is not a copy of the first. It has to re-hit your sealed standard on its own machines, and the first run off a new line almost always varies more than a settled one — a slightly different hand, a seam that behaves differently, a colour that needs another dip to land. That is normal, and it is exactly why a first run through a new partner earns tighter inspection, not less.
The clock is the other cost. Requalifying runs on ordinary sample timelines — for stock constructions a sample lands in around three to four days, and a custom first sample is closer to a week — and then bulk queues behind that, plus the extra QC a first run deserves. Rerouting is far faster than starting a sourcing search from zero, but it is measured in days and weeks, not hours. Anyone who tells you a switch is seamless is selling the idea, not doing the work.
Which is the real case for a network: the slow, expensive part is the qualification, and you want it done before the emergency rather than during it. A second specialist approved in a calm month is insurance you can draw on in days. The same qualification attempted the week your only factory goes down is a scramble — the worst possible conditions under which to hold a new supplier to your standard for the first time.
When one factory is the right call
A network is not the right answer for everyone, and pretending otherwise would be the same loose over-claiming this guide argues against. Plenty of brands are well served by a single factory, and adding qualified alternatives they will never use is just cost and coordination they do not need. The goal is to choose the dependence deliberately — not to maximise the number of suppliers on a slide.
Three situations genuinely favour one factory. The first is a single simple, stable product: one construction with no seasonal reinvention, where the case for a second qualified line is weak. The second is a deep, proven relationship in which you are a valued account rather than a marginal one — years of clean orders buy you priority and an honest early warning that a new brand ordering near a minimum cannot assume. The third is volume: if your orders fill a meaningful share of a factory’s calendar, you are the client others get rescheduled behind, and much of the capacity risk described here is simply not aimed at you.
Even then, the useful question is whether your dependence is deliberate and hedged or accidental and total. A brand that has consciously chosen one factory, knows what its fallback would be, and keeps its sealed reference and patterns in its own hands is in a very different position from one that drifted into single-sourcing and would be starting from nothing the day that factory closed. Single-factory can be exactly right. Single-factory with no idea what you would do next rarely is.
What to have in place before you need it
Continuity is bought before it is needed, so the questions are worth asking while things are calm. For each construction you depend on: is there a second specialist that could be qualified for it, and how long would that take from a standing start? Who physically holds the sealed reference and the patterns — you, a coordinating partner, or only the factory? Is there a fabric you could approve in parallel, and has anyone actually tested it? And what, precisely, is the plan if this factory is full straight through your peak?
The things worth holding are unglamorous and cheap against the price of a missed season: your own copies of the sealed reference, specification and patterns, so the standard stays portable; at least one construction-appropriate alternative identified, even if not yet fully qualified; a fabric fallback for anything critical; and an agreed, plain-language trigger for the point at which a slipping date becomes a reroute rather than one more week of waiting. Keeping these live is the coordinating partner’s job, so that a failure at one facility is a substitution and not a stoppage — moving your order to another qualified partner, not adding a line we own.
You cannot control whether a factory has a full quarter, a failed audit or a bad year. That risk is real, and it is not yours to remove. What you can control is whether that factory’s bad quarter has to become your brand’s bad quarter too — and that is decided long before the problem arrives, by whether your program was ever allowed to rest on a single point of failure in the first place.