Every ribbon programme has a document that fixes the product: the purchase specification. Far fewer have a document that fixes the process — how the factory turns that specification into metres, on which line, in which week, with which lot. That document is the control plan, and the gap between the two is where most late deliveries, most lot-to-lot inconsistency, and most capacity arguments live.
This guide is about the production side of OEM ribbon buying: what a control plan should contain, how a factory allocates capacity when it is oversubscribed, how to read a capacity claim honestly, what changeover actually costs you in lead time, how to size lots that the line can schedule, and how the Lunar New Year shutdown lands on every seasonal plan. It is written for brand buyers, product managers and supply-chain planners who have to hit a date and inherit the quality of what arrives.
A purchase order answers a commercial question: what did you agree to buy, at what price, by when. It says nothing about how the factory will get there.
That gap matters because ribbon production has a specific characteristic that makes it different from most made-to-order goods. A factory running narrow-width ribbon has a fixed number of lines, and each line can run one construction at a time. Your order is not a job that gets scheduled into a queue of equivalent work — it is a claim on a specific piece of equipment for a specific number of hours, competing with every other customer who wants the same week.
Three consequences follow, and they drive everything in this guide.
Capacity is finite and it is allocated, not sold. When a factory is oversubscribed in peak season, something has to give. The default mechanism in almost every factory is first-come-first-served, which means your date is protected by the date you happened to order, not by the importance of your programme.
The number of SKUs you order changes your lead time. Every new colour and every new construction consumes line time in changeover, and changeover time produces no saleable ribbon. A factory with fifteen active constructions across four lines is spending a meaningful share of its schedule not making anything.
Calendar risk sits upstream of the factory. The shutdowns, holidays and maintenance windows that determine whether you receive goods in March are largely outside your factory's control and largely inside its calendar. If you plan a spring range to start production in late February, you are planning into the single most predictable delay in the Chinese manufacturing year.
The good news is that all three of these are manageable with documents rather than luck. That is what a control plan is for.
A control plan is the document that says: this is how we will produce your specification, here is where we will check it, here is what happens when the check fails. Its purpose is to make the production process explicit enough that it does not depend on which shift, which operator or which week it runs.
For ribbon, a control plan should cover at least nine stages. We use this list when we write one for a customer, and we think it is the right skeleton for a buyer to ask for:
| Stage | What must be controlled | Typical check | Defect it catches |
|---|---|---|---|
| 1. Yarn / raw material receipt | Fibre type, count, denier, lot identity | Incoming ticket check against spec | Wrong material entering the programme |
| 2. Warping / preparation | Number of ends, tension, pattern | Pattern read against the construction sheet | Wrong weave structure |
| 3. Weaving | Width tolerance, construction, pick density, selvage condition | Width measured at three points; visual edge check | Width drift, yarn pull-out, open selvage |
| 4. Dyeing | Shade, dye lot, temperature and time, batch size | Colour-difference reading against the approved reference | Shade drift, batch inconsistency, colour contamination |
| 5. Colour matching / lab dip | Conformance to the approved sample | Standard light box comparison | Wrong shade shipping |
| 6. Finishing | Starch, softness, calendering, heat setting, pile direction for velvet | Hand and stiffness check; rub test where specified | Wrong hand, glazing, crushed pile |
| 7. Decoration | Print colour sequence, registration, coverage, adhesion, foil temperature | Registration check; rub test on printed area | Mis-registration, poor adhesion, colour mismatch to print artwork |
| 8. Slitting and winding | Cut width, roll length, core size, direction of winding, edge cleanliness | Width and length verification; roll label check | Wrong width, wrong core, ribbon wound against the grain |
| 9. Packing and labelling | Carton quantity, labels, lot number, shipping marks | Packing list against release note | Mixed lots, missing traceability, wrong carton count |
Two design principles matter more than the specific parameters.
Each stage needs a reaction plan, not just a limit. A control plan that says "width 19 mm ±0.5 mm" but says nothing about what happens when a roll measures 20 mm is half a document. A useful control plan states the reaction: hold the lot, identify the last conforming roll, re-inspect back to the last good check point, and notify the buyer. Without a reaction plan, a deviation is simply recorded and production continues.
Each stage needs a named owner and a sampling rule. Who checks, at what frequency, on what sample size. "Quality checks" is not an instruction. "Weaving, operator, every roll, width at head/middle/tail, record on the roll ticket" is.
For a buyer, the value of the control plan is indirect but real: you are not going to write it and you are not going to audit it line by line. What it tells you is whether the factory has thought about how your specification will actually be produced, or whether your specification arrived and was absorbed into a general process. Factories that write real control plans tend to be transparent about the parts they are less confident about, which is itself the most valuable information you will get.
When capacity is tight, something has to be allocated. Understanding the default mechanism tells you how much of your date you actually control.
Most factories allocate in a fixed priority order, and it is rarely "the customer who pays most". It is usually a combination of:
Notice what is not in that list: the commercial value of your programme to the end customer, or the criticality of your launch. Those factors only enter the allocation if you put them there. Which is the argument for the next section.
Every factory has a headline capacity number. Most of those numbers are real, and most of them are also misleading, because they are stated as a peak rather than as an average.
Here is the reconstruction we recommend when a factory quotes a capacity figure. Start from the stated headline and discount it four times.
Deduction one: shift pattern reality. A line running three shifts produces more than the same line running one, but the third shift on weaving and winding is usually less efficient and often staffed by fewer experienced operators. If a factory assumes a full three-shift yield on every line, discount the night shift output by a realistic proportion.
Deduction two: changeover time. This is the deduction buyers most often skip, and it is usually the largest. If a factory runs thirty constructions a month across four lines, it is performing many changeovers, and each one consumes line time that produces nothing. Ask directly for the average changeover per construction family and subtract it.
Deduction three: maintenance and shutdown. Loom maintenance, winding downtime, dyeing tank cleaning, roller replacement, plus the annual Lunar New Year break. In a peak year this can remove four to six weeks of effective capacity from a twelve-month calendar. The daily figure quoted in a brochure usually assumes a full working calendar.
Deduction four: yield loss. A factory running realistic AQL inspection will ship a percentage of lots short, rework a further percentage, and scrap a further percentage. Effective shipped capacity is below produced capacity by the yield rate, and in seasonal colours that loss is higher because the shade tolerance is tighter and the colour matching consumes more test metres.
The test of an honest capacity claim is simple: ask for three consecutive months of actual monthly output by construction, and rebuild the annual number from what actually happened rather than what the brochure says. A factory that gives you that breakdown without argument is telling you the truth. A factory that offers only the headline figure is offering you a marketing number, and you should plan your orders as though it were 20 to 30 per cent higher than reality.
One more piece of evidence worth requesting: the ratio of distinct constructions to lines per month. If a factory runs four lines and the answer is "about six constructions per line per month", the changeover load is heavy and lead times will be long. If the answer is "one to two", the schedule is clean and dates are reliable. This single ratio often tells you more about delivery reliability than any certificate in the file.
This is the single most useful piece of production economics in ribbon, and almost no buyer asks about it.
Changeover is everything that happens on a line between finishing one construction and starting the next. On narrow-width ribbon it typically includes:
The duration depends on how much changes at once. A colour change on the same construction might take a few hours. A change of colour, width and construction simultaneously can take a full working day or more. And it can only be parallelised across lines if the next construction is compatible with that line — a narrow-dress line cannot pick up a wide-elastic job, however idle it looks.
Three commercial consequences follow, and each of them should be reflected in how you order.
Your minimum order quantity is a changeover artefact. When a factory asks for 500 metres or 1,000 metres, the real reason is usually that below that quantity the changeover consumes a disproportionate share of the run. The factory is not being difficult; it is protecting its own schedule. Where you have volume, ask whether a larger single order across several colours can be grouped on one line set — that is, ask whether they can change colour only, rather than colour and construction together. Grouping across colours on a shared construction is one of the cheapest lead-time improvements available, and it is rarely asked for.
Your lead time grows with your SKU count. A brand with sixty active ribbon SKUs is structurally slower than a brand with twenty, because the factory's schedule is fragmented. If your own range is drifting towards many small variants, you are paying for fragmentation at both ends.
Your delivery date is a function of position in the queue, not order size. Once the factory is committed to a changeover sequence, the order that arrives mid-sequence waits for the whole sequence to finish. Ordering sooner and ordering more of fewer items both help; ordering a tiny extra run in the middle of someone else's sequence does not.
Buyers tend to size orders by what is comfortable for their own inventory position. That is the wrong starting point, because the order also has to be schedulable by the factory, and those two constraints rarely coincide.
A better approach: size the lot against the changeover economics and your own demand curve, and then split it.
One lot per colour per construction. This is the starting point. A lot that mixes colours or mixes constructions is a lot with a changeover inside it, which means the run is split, the inspection is split, the traceability is split, and the lead time is longer than the two runs added together would suggest.
Size the lot so it consumes a meaningful share of a line-day. If a colourway runs two hours of line time, ordering one thousand metres when the run needs eight is not efficient use of a slot; the factory will slot it around bigger work and your delivery date becomes probabilistic. Either increase the lot to fill the slot or decrease the number of colours in the order so each one gets its own slot.
Split seasonal programmes into a base lot and a top-up lot. This is where most seasonal delay actually originates. A brand that places one large spring order in January for goods needed in June is exposed to the Lunar New Year shutdown and to a February queue. A brand that places a base lot early — the quantity needed for the first production run or the first wave of retail delivery — and a top-up lot later has converted one risky large bet into two smaller, better-timed ones. The base lot protects the launch. The top-up lot absorbs the demand you did not forecast, and it can be scheduled around a colour you have already validated.
Tell the factory your lot pattern in the specification. This is the least-effort improvement available to a buyer. If the factory knows that you will take eight colours in the same construction in the same month, it can plan the changeover sequence to change colour only, run them back to back on the same line, and finish your programme in a fraction of the time it would take switching construction eight times. Grouping your own orders is free, and it is worth more than most of the negotiating points buyers spend their time on.
Keep a live colour and construction map. A single page showing which colours run on which construction, in which month, with which lot sizes. It stops the range drifting, it lets you spot the two SKUs that could share a setup, and it gives the factory a document it can actually schedule against. When your programme reaches the stage where the same construction appears in four colours across two months, that map is the difference between a grouped run and four separate changeovers.
Every factory has planned downtime. Some of it is visible on a calendar you can ask for; some of it is not planned at all, which is the more dangerous category.
The planned items worth asking about are line-level: loom and winding maintenance windows, dyeing tank cleaning cycles, calender and finishing roller replacement, and the annual factory shutdown around the Lunar New Year.
That last one deserves specific attention because it is the most predictable and most damaging delay in the Chinese manufacturing year. The Lunar New Year period typically removes two to three weeks of effective production from the calendar, plus travel days on either side that have historically cost another week of output. A factory that closes for three weeks of the public holiday, and whose staff have travelled and returned across a much wider period, is effectively out of production for roughly a month in February.
Three practical rules for anyone building a spring or summer range:
It is worth noting that the shutdown is not a failure of the factory, and a factory that tries to work through it at full rate will deliver you goods that were made by exhausted staff on unfamiliar shifts, which is a worse outcome than a later date. Treat the calendar as a fact to plan around, not an obstacle to argue about. When you ask a factory for its shutdown calendar, you are doing the sensible thing, and the answer should come quickly.
Everything above leads to one practical conclusion. If you have a launch date, a seasonal peak, or a retail window that cannot move, you needa booked production slot, not just a purchase order. A purchase order tells the factory what you want and when you want it. A booked slot tells the factory that a specific line, in a specific week, is reserved for your programme and that your work comes before other work scheduled into that week. The difference between the two is the difference between being in the queue and being in the schedule. What a booking usually contains:
None of this requires you to manage a factory. It requires six fields in your ribbon specification and one internal routine.
Field one: the control plan. State that the factory shall maintain a written control plan for your SKU, covering the nine stages listed above, and shall make it available on request. Do not ask them to send it to you routinely — it will change, and you will end up approving a version you did not read. Ask to see the current version once, at the start, and once again at your first reorder.
Field two: the reaction plan. For each controlled parameter, require that an out-of-tolerance result places the affected lot on hold, identifies the last conforming production point, and triggers a written notification to you within a stated number of hours. This is the field that converts a control plan from a record into a control.
Field three: the capacity commitment. State the annual and peak-month capacity the factory is committing to for your construction, the basis on which it is calculated, and the maximum ratio of your construction to the factory's total load during peak. A factory that is willing to state a number, and a basis, has thought about whether your programme fits.
Field four: the booking mechanism. State the reserved-slot terms: the notice period for booking, the minimum quantity for a booking, the deposit structure, and the deadline for confirming the balance.
Field five: the lot and grouping rule. State your expected lot pattern — one colour per lot, grouping across colours on shared construction, base lot plus top-up lot for seasonal programmes — and require the factory to schedule your work to minimise changeover. This clause is what turns your own order discipline into a factory scheduling instruction.
Field six: the calendar commitment. Require the factory to publish its planned maintenance and shutdown calendar for the next twelve months, to notify you of any unplanned line downtime exceeding a stated duration, and to confirm that your booked weeks are unaffected by an unplanned stoppage or are re-booked within a stated number of days.
The internal routine is a quarterly production review, and it should be short. Four questions, each answerable in one line:
Those four numbers, reviewed four times a year, will tell you more about the reliability of your ribbon supply than a factory audit ever will. And they will tell you early, which is the only time the information is useful.
If you are qualifying a supplier, these seven questions reveal whether production is under control, without a site visit:
A factory that answers all seven with numbers and documents is describing a controlled process. A factory that answers with adjectives — “flexible”, “fast”, “we have plenty of capacity” — is describing an aspiration. Notice that none of these questions are adversarial. They are all reasonable planning questions that any serious buyer asks, and a factory that treats them as reasonable is telling you something useful about how it will behave when your programme has a problem.
None of this makes the ribbon better, and none of it lowers the unit price. What it does is convert your delivery date from something the factory decides into something you have arranged in advance.
Consider what production control buys you in three situations. When the factory is oversubscribed in November, you have a booked slot, so your goods are in the schedule rather than in the queue. When the factory changes your construction because it improved their own line, you have a control plan and a re-approval trigger, so the change arrives as a question rather than as a surprise in a carton. And when the Lunar New Year shutdown lands on your plan, you have already placed the base lot and booked the restart week, so it is a known quantity in your schedule instead of an explanation you are writing to a customer in March.
There is a fourth effect that is harder to measure but consistently matters. A brand that can promise a delivery date, and hit it, is a brand that its customers plan around. In a category where the product is visually trivial and the switching cost to the customer is almost zero, delivery reliability is one of the few things that genuinely distinguishes one supplier from another. Most of your competitors are negotiating on price. Not many are asking for the changeover ratio.
Our own production is run this way. We maintain a written control plan for every custom programme, plan capacity by construction family, group colours onto shared setups wherever the construction allows it, and publish our maintenance and shutdown calendar to customers booking ahead. Our factory floor runs more than 200 people at a daily output above 100,000 metres across our own 15,000 m² site in Xiamen, and the reason a booked week means something with us is that we hold the slot rather than re-sorting the week by whoever ordered first.
Smith Ribbon has been manufacturing ribbon, bows, tassels and decorative trim in Xiamen since 2004, with our own 15,000 m² factory, more than 200 employees, daily output of 100,000 metres, and certification including OEKO-TEX, FSC, BSCI, SEDEX/SMETA and ISO 9001. We ship to more than 50 countries and support OEM and ODM customisation from 1,000 metres, with small-batch runs from 500 metres.
We run a written control plan for every custom programme, plan capacity by construction family, group colours onto shared setups where the construction allows, and publish our planned maintenance and shutdown calendar to customers booking ahead. Whether you need a control plan written into your specification, a reserved production slot for a launch date, a lot pattern that minimises your changeover count, or a costed quotation for a booked production programme, send us your specification and we will tell you plainly what we can commit to and where the constraint sits.
Email: xmmsd@126.com | Mobile / WeChat: +86 137 7995 1780 | Office: +86-592-5095373
These are the questions brand buyers, product managers and supply-chain planners most often ask us about production control in ribbon programmes: what a control plan should contain, how to verify a real capacity figure, what changeover actually costs, how to size lots the line can schedule, how the Lunar New Year shutdown affects delivery dates, and how capacity booking differs from a purchase order. If your question is not here, ask us directly.
A ribbon control plan is the document that translates your purchase specification into the process steps on the factory floor. It names, for each stage from yarn preparation through weaving, dyeing, finishing, slitting and packing: the machine or line family, the process parameter that must be held, the inspection point and the sampling rule, the defect class it catches, and the reaction plan when the parameter moves out of tolerance. It should also state who signs each stage and what triggers a hold. For a buyer, the value is not in owning it, but in asking to see it: a factory that can show you a control plan for your SKU has thought about how your specification will actually be produced.
Work in metres per line-hour rather than daily headline output, and then discount it. A factory claiming 100,000 metres a day usually means a peak figure at full line running with the best-performing construction. Rebuild the number yourself: metres per hour by construction type, multiplied by the number of lines that can actually run that construction, multiplied by realistic shift patterns, minus changeover time, minus planned maintenance, minus the yield loss that realistic AQL inspection implies. A capacity figure that has no changeover deduction, no maintenance allowance and no yield deduction is a sales figure, not a planning figure. Ask for three consecutive months of actual monthly output by construction and compare it with the claim.
Changeover is everything a line must do between finishing one construction and starting the next: clearing the previous colour, threading new yarn or warp, changing width or reed, cleaning the dye machine and tanks, re-setting the print screens, and running off the first metres. On narrow-width ribbon, a changeover can run from a few hours for a similar construction to a full day when the colour, width and construction all change at once. Because each changeover consumes line time that produces no saleable output, a factory running many small SKUs can look busy while delivering very little. This is the mechanism behind most minimum order quantity and most quoted lead time, so ask a factory what its average changeover is per construction family before you discuss MOQ.
Plan in whole production lots sized against the factory's changeover economics and against your own demand, not in arbitrary round numbers. A good rule is one lot per colour per construction, with each lot large enough that the run consumes a meaningful share of a line-day rather than a fraction of it, and small enough that you are not funding months of inventory in colours you may not re-order. For seasonal programmes, split the quantity into a base lot and a top-up lot rather than one large order, and book the base lot's production slot early. Tell the factory your planned lot pattern in the specification so the line can group the same construction and reduce your changeover count, which usually shortens your lead time without any price negotiation.
They affect them more than most buyers expect, because the maintenance that matters is rarely the factory-wide holiday. It is loom and winding downtime, dyeing tank cleaning, calender and finishing roller replacement, and the annual shutdown around Chinese New Year. In our industry the Lunar New Year period typically removes two to three weeks of effective production from the calendar, and it lands at the worst possible point for anyone building a spring or summer range. Ask a factory for its planned maintenance and shutdown calendar for the next twelve months, ask specifically about Chinese New Year dates, and place your forward orders so that the goods are finished and shipped before the break rather than scheduled to start after it. This is the single most common cause of an avoidable late delivery in Chinese ribbon sourcing.
By default, your order goes to the back of the queue, because capacity allocation in most factories is first-come-first-served unless it has been reserved in advance. That is why capacity booking, not the purchase order, is the document that protects your date. A booked slot commits the factory to reserve specific lines and specific weeks for your programme, usually against a deposit or a minimum volume commitment, and it survives the reordering of other customers' work. If your programme has a fixed launch date or a seasonal peak, a booked slot with a named week is worth considerably more to you than a price concession, and it is usually cheaper to obtain.
Use three indirect tests. First, ask for the monthly output by construction for the last six months and check whether the totals are consistent with the claimed daily capacity once you allow for changeovers, maintenance and shutdown. A factory claiming very high capacity but showing a long tail of small monthly runs is telling you that changeover is eating its schedule. Second, ask how many distinct constructions ran on the line each month. A number close to the number of lines suggests poor grouping and long lead times. Third, ask for a recent example of a rush order that was delivered on time and how it was accommodated, then ask what it displaced. A factory that answers all three honestly, with numbers, is describing a controlled process.