Every ribbon buyer eventually learns the same lesson: a quotation is a position, not a fact. Two buyers can receive materially different quotes for an identical specification, and both can be internally coherent, because the difference lives in assumptions that were never made explicit. A should-cost model is the discipline of building the cost of the product yourself so you can read a quote against your own number instead of against the last quote. It is not a negotiation weapon. It is a way to know whether a price is reasonable, and when a deviation is worth asking about.
Every ribbon buyer eventually learns the same lesson. A quotation is a position, not a fact. It is the supplier's account of how your order will be built, priced and delivered, expressed in a number, and adjusted according to how much the buyer believes the supplier needs the order. Two buyers can receive materially different quotes for an identical specification and construction, and both quotes can be internally coherent. The difference lives in assumptions the buyer did not make explicit: which cost elements the supplier bundled, which they forgot, which they assumed, and which they left out because a less experienced buyer would not have noticed.
A should-cost model is simply the exercise of building the cost of the product yourself, from first principles, so that you can read a quote against your own number instead of against the last quote. It is not a negotiation weapon in itself, and it is not a substitute for a supplier relationship. It is a discipline that tells you whether a price is reasonable, and when a deviation is worth asking about.
This guide covers how to build a should-cost model for ribbon and decorative trim: how to structure it, what the cost layers actually are, which assumptions drive the number most, how to bridge from ex-works to landed cost, and how to use the model without damaging a supplier relationship that is working.
A useful model is not one enormous spreadsheet. It is a short chain of quantities and rates, each of which you can interrogate, ending in a cost per unit and a cost per metre that you can compare against a quote. The chain looks like this.
Quantity build. Start from the finished unit, which for most ribbon programmes is a roll of a stated length, or a metre of ribbon, or a finished bow. Work backwards: how many metres of ribbon does one bow consume, including the cutting and tying waste? How many linear metres of greige are needed to yield one linear metre of finished ribbon, accounting for the width loss between the loom width and the finished ribbon width? How many picks and ends does that represent in weaving terms? The quantity build is where most of the genuine engineering savings hide, and it is also where the most quoted-versus-actual gaps appear.
Material layers. Yarn cost is the largest single element in most programmes and the one most often quoted as an index rather than as a number. Build it as fibre price, converted into yarn cost per kilogram, converted into metres per kilogram using the pick density and yarn count, then converted into cost per metre of ribbon. When you do this properly, you discover that the yarn price the supplier quotes is rarely the yarn price they paid. There is a conversion cost in between, and a quality and origin differential, and sometimes a specific yarn blend that costs materially more than the generic market reference.
Conversion layers. These are the operations that turn greige into finished ribbon, and each has a real rate and a real yield loss. Weaving or fabric formation. Warp and weft preparation. Sizing. Dyeing, either yarn-dyed or piece-dyed, with the associated dye-house consumption, water and energy. Printing, either yarn-printed, rotary-printed, or digitally printed, with setup and cylinder or screen costs amortised across the run. Finishing: heat setting, calendaring, brushing, flocking if velvet, metallic lamination or hot stamping. Slitting to width. Winding and packing. For a bow programme, add the conversion: cut lengths, folding, hand-tying or pre-tie assembly, tagging, bundling.
Overhead and margin. Fixed manufacturing overhead absorbed at a rate consistent with the volume you have committed to. Then the margin. The buyer's model should express margin as a rate or a percentage, because that is how most suppliers actually build it, and a model that expresses margin as an absolute number per order just obscures the mechanism.
That gives you a cost per unit at mill gate. Everything after that is logistics, duty and the working cost of capital, and those are the subject of a separate section, because they are the parts most often quoted on a different basis between buyers.
A model with fifty inputs is only useful if you know which five matter. For ribbon, the sensitivity ranking is fairly stable.
| Driver | Typical influence | Why it dominates |
|---|---|---|
| Yarn price and yarn count | Highest | Usually the largest single cost element, and the one most exposed to market movement |
| Pick density and width | High | Determines metres per kilogram and therefore yarn consumption per finished metre |
| Colourway complexity | High | Dye lot minimums, lab dip approval cycles and the number of dyeings in the programme |
| Print setup amortisation | High at low volume | Cylinders, screens, digital setup and approval samples are fixed costs spread over the run |
| Order quantity and run length | High | Drives setup amortisation, dye lot efficiency and changeover losses |
| Conversion rate per operation | Medium | Less volatile than material, but it compounds across many operations |
| Waste and yield | Medium | Width loss, cutting waste, and the difference between nominal and actual usable length |
| Margin and overhead | Medium | Lower visibility, higher sensitivity to the buyer's negotiating behaviour |
Two consequences follow. First, a negotiation focused on the visible price per metre is usually targeting the wrong variable, because the same per-metre price can hide very different cost structures. Second, when a supplier's quote moves unexpectedly, the reason is almost always in one of the top four rows: the yarn price moved, the density changed, the colourway got more complex, or the order got smaller and the setup spread over fewer metres.
This gives a buyer a much better question than "why did the price go up". The better question is "which of these four changed, and what is the corrected comparison". It is a factual question, it is easy for a supplier to answer, and it moves the conversation from the price to the specification, which is where it belongs.
When a quote arrives, build the model and compare. The comparison is more informative if you compare structure rather than total. For each layer, note what the supplier appears to have assumed and what you assumed. Then classify the differences.
Explicable differences are the good ones. The supplier may quote a higher yarn price because they are using a specific mill's yarn with a specific certification, and you modelled a market index. The supplier may include a setup amortisation over a shorter run because their minimum run is longer than yours. The supplier may include a slitting or winding operation you folded into the weaving rate. These are not disagreements; they are two valid constructions, and the resolution is a specification clarification that often improves the programme.
Scope differences are the ones that cause disputes later. The quote may exclude a finishing operation you assumed was included. It may exclude the hang tag, the box, the tissue, the drop ship, or the retail-ready packing. It may include a cost that is really part of a component you buy elsewhere. Writing the scope boundary down at quote stage is the single most effective way to prevent the argument resurfacing at delivery.
Unexplained gaps are where margin and risk premium live. A quoted price that sits well above the model with no discernible structural reason is not necessarily wrong, but it is a legitimate question. Ask what the number contains. In my experience, most of these resolve into one of three things: a risk premium for a new programme with no history, a margin cushion the supplier holds because it expects the quote to be negotiated down, or a genuine unrecognised cost the supplier has not modelled because it has never had to explain one. Each of those has a different correct response, and only the first is a reason to keep the number as is.
Ex-works quotes are not comparable across suppliers unless you normalise the delivery basis and the inclusions. The bridge has four components.
Freight and insurance. Ribbon is volumetric. Chargeable weight for air freight is the greater of actual gross weight and volumetric weight computed from carton volume, which for most freight conventions is 166 kg per cubic metre. A consignment that is 40 percent denser than the volumetric allowance will be billed on the volumetric basis, and the difference can be a factor of one and a half or more. For sea freight the basis is volume in CBM, and container fill rate is the dominant lever. This is where the roll length and carton geometry decisions from the purchasing specification pay off, and it is why a ribbon quote that looks competitive ex-works can be uncompetitive on arrival.
Duty and trade measures. The classification of woven narrow fabric determines the duty treatment in each destination market, and the classification of a jacquard woven ribbon is not always the same as that of a printed ribbon even when both are described as ribbon. Model duty as a percentage of the customs value rather than assuming it, and note that the preferential origin of the material matters as much as the origin of the conversion.
Origin handling. A DDP quote bundles freight, duty and clearance into one number, which is convenient for the buyer and hides where the cost sits. For a should-cost model, always ask for the ex-works component and build the rest yourself. Two DDP quotes from different suppliers can differ substantially in the underlying ex-works price, and knowing which one is cheaper is useful even if you end up purchasing on DDP.
Working capital and the cost of a delay. The final bridge item buyers most often omit. Goods in transit are capital that has left the balance sheet and not yet become sellable inventory. A sea shipment has a long in-transit period and a risk of delay that is not zero; a slower mode with a lower freight cost is not automatically the cheaper option once the cost of a stockout or a missed launch date is included. Model the delay cost explicitly rather than leaving it in the intuition of the person choosing the mode.
A should-cost model is a diagnostic tool. How you deploy it determines whether it improves your supply base or destabilises it.
Do the modelling quietly first. Your model is an internal instrument. Opening it in the first meeting with a supplier you have not yet worked with reads as an attempt to prove the supplier is overcharging, and it establishes the relationship on adversarial terms before either side has exchanged anything.
Use it to test the specification, not the margin. The most valuable use of a model is often upstream of price: to check whether the specification you are about to release is actually the cheapest way to achieve the visual result you want. Widening a ribbon, adding a colourway, requesting a print setup or upgrading to a certification all have cost consequences that are easier to evaluate before the brief is written than after the quote is received.
Use it to explain, not to squeeze. When a quote and a model diverge, the productive conversation is a joint walk through the two constructions: this is the yarn assumption, this is the setup amortisation, this is the conversion rate. Most divergences resolve into a shared understanding, and the supplier ends up with a clearer brief and you end up with a quote that is closer to its own build-up. That outcome is durable. A quote reduced by pressure and not by understanding will drift back up at the next order, and the supplier will learn that your targets are not real.
Accept that a margin is a legitimate part of the price. A model that concludes the fair price is materially below the quote is a finding, not a discovery of wrongdoing. The mill has overhead, risk, financing cost and a return obligation. If the modelled cost leaves nothing, the programme is not commercially durable, and the resulting fragility will show up as a quality or delivery failure two orders later. Several established ribbon mills operate on margins that look thin in isolation and are perfectly reasonable once overhead absorption and the cost of capital are modelled properly.
For a brand buyer building a model for the first time, the sequence below works and does not take long.
Start with a single high-volume, well-specified SKU that you have actually received and inspected. Using a real product means the quantity build is verifiable rather than theoretical, and you can compare the model against both the quote and the goods.
Build the quantity chain first and check it against the physical product: how many metres did one bow actually consume, how much width was lost between loom and finished width. Do this before any pricing work, because a five percent error in the quantity build is a five percent error in everything downstream, and no amount of rate precision recovers it.
Then build the material and conversion layers with rates you can defend, and mark clearly which rates you have verified from the supplier versus which you have assumed. The assumption log is more useful than the model, because it tells you exactly which questions to ask.
Next, add the landed cost bridge and convert to a delivered unit cost. Compare against the quote, layer by layer, and classify the differences. Expect two or three genuine scope disagreements on the first pass, because the specification was never written to answer them.
Finally, decide what the model is for. If it is for a single supplier's ongoing negotiation, keep it narrow. If it is for comparing suppliers, it must be built on an identical specification basis for all of them, which usually means writing a tighter brief than you have been working with. That tightening is itself valuable, because it is the same document that reduces the number of questions a quote generates.
The value of a should-cost model is not that it produces a number you can use to beat a supplier. It is that it changes the quality of your specifications and the quality of your questions. A buyer who can construct the cost of their own product stops reacting to price movements and starts understanding them. That is a permanent reduction in the cost of the relationship, because the two people on either side of the table are finally working from the same structure, and the argument moves from "your price is too high" to "these two assumptions differ, and here is which one is right".
It is a build-up of the cost of your own product from first principles: the quantity chain, the material layers, the conversion operations, overhead and margin, and then the landed cost bridge. Its purpose is not to produce a number you can use to beat a supplier. It is to change the quality of your specifications and the quality of your questions, so that a price movement becomes explainable rather than something you react to.
Start from the finished unit and work backwards: how many metres one bow actually consumes including cutting and tying waste, how many linear metres of greige are needed to yield one linear metre of finished ribbon after the width loss between loom width and finished width, and how many picks and ends that represents. Do this before any pricing work, because a five percent error in the quantity build is a five percent error in everything downstream.
Four dominate: yarn price and yarn count, which are usually the largest single element and the one most often quoted as an index rather than a number; pick density and width, which determine metres per kilogram; colourway complexity, which drives dye lot minimums and lab dip cycles; and order quantity and run length, which drive setup amortisation and dye lot efficiency. When a quote moves unexpectedly, the reason is almost always in one of those four.
Compare structure rather than total. For each layer, note what the supplier assumed and what you assumed, then classify the difference as explicable, which usually reflects a legitimate difference in construction and often improves the programme; as scope, which usually reflects an unwritten boundary such as an excluded finishing operation or an excluded hang tag; or as unexplained, which usually turns out to be a risk premium for a new programme, a negotiation cushion, or a real unmodelled cost.
Four components. Freight and insurance, remembering that ribbon is volumetric and air freight is billed on the greater of actual gross weight and carton volume at 166 kg per cubic metre, while sea freight is billed on CBM and container fill rate is the dominant lever. Duty and trade measures, modelled as a percentage of customs value with attention to whether the classification of a jacquard woven ribbon differs from a printed ribbon. Origin handling, where you should ask for the ex-works component and build DDP yourself. And working capital plus the cost of a delay.
Do the modelling quietly before the first meeting, because opening a model in the first conversation with an untested supplier reads as an attempt to prove overcharging. Use it to test the specification rather than to squeeze the margin. Use it to explain rather than to squeeze, so divergences resolve into a shared understanding. And accept that a margin is a legitimate part of the price: a model that leaves nothing is not a better buy, it is a fragile programme that shows up later as a quality or delivery failure.
Send us one SKU you have actually received and inspected. We will build the quantity chain with you, show which cost layers our own production data supports, and flag the questions we would expect a buyer to ask before the brief is released.
Talk to a compliance engineer: +86 13779951780 (WeChat) | xmmsd@126.com | www.smithribbon.com
© 2026 Xiamen Smith Ribbon & Bow Co., Ltd. | Professional Ribbon Manufacturer Since 2004
Smith Ribbon OEM since 2004 — woven ribbon, satin, grosgrain, velvet, organza, RPET, printed bows, decorative trims, and private-label accessories for global brand buyers.