---
title: "Against made to order as a first strategy"
description: "Made to order defers the fit, supplier and pricing decisions a first range exists to force, while production economics and lead times work against a very small run."
url: "https://surajmalstudio.com/insights/2026/08/against-made-to-order-as-a-first-strategy"
---
# Against made to order as a first strategy
14 min read Costs, orders and delivery Published 29 Aug 2026
Made to order defers the fit, supplier and pricing decisions a first range exists to force, while production economics and lead times work against a very small run.
Making after an order changes when production happens. Approved fit, available materials, repeatable making and a credible dispatch promise still need to be settled before selling.
Made to order relocates the risk in a first range. It does not remove it. It can reduce unsold finished stock, while leaving commitments in materials and development and introducing a longer customer wait. Whether that trade works depends on the product, maker and readiness of the fit programme.
## 01For some products, working in ones is the product
The model deserves a fair hearing before the argument against it. Cutting only what has been paid for is a real advantage, and it is the advantage most first ranges are told to want. Nothing is written off. Nothing is discounted to clear a size that never moved. Cash converts in the right direction, from customer to maker, instead of sitting on a shelf waiting to be proved wrong. For a small number of products this is simply the correct answer and stays the correct answer: garments cut to a customer’s own measurements, pieces built around a chosen component, work where the variation is the product rather than a complication of it.
The conditions that make it work can be listed, and they are specific. The pattern is settled, so no garment is teaching the brand something a size set would have taught sooner. The maker [has already produced the style in quantity and can repeat it](https://surajmalstudio.com/insights/2026/07/the-second-order-is-the-real-test), so a single unit is a repetition rather than a first attempt. The cloth is either held by the brand or sold by a supplier in the amount one garment needs. The customer has a reason to wait that the garment itself supplies, whether that is their own measurements, a cloth they chose, a monogram or a component they picked. And the retail price can carry the cost of working in ones, because the variation is what is being paid for. Where those hold, made to order is not a compromise made for cash reasons. It is the design.
There is a narrower use inside an ordinary produced range, and it is also legitimate: an extended size sitting outside the graded run, a repeat of a colour that sold out while the cloth is still with the supplier, a style offered in a made-to-measure version alongside the standard one. Each of those sits on top of an operation that already works. None of them is the operation.
The problem is not the model. It is the claim made for it, that this is the low risk way to begin. Risk in a first range is not only the risk of unsold clothes. It is the risk of reaching the second year without knowing whether the fit works, whether the maker can repeat the garment, or which part of the range the customer actually came for.
## 02Setup is recovered on the run, not on the piece
Start with [what production actually costs](https://surajmalstudio.com/guides/what-drives-development-cost) . A factory recovers its setup on the run, not on the piece. Marker making, cutting, machine setting, thread and needle changes, trims arriving in usable amounts, the first pieces checked off the line: that work happens once whether the order is small or large, and a very short run carries all of it. Fabric behaves the same way. Cloth and colour are production operations at the mill, so [buying in small amounts costs more per metre when it is available at all](https://surajmalstudio.com/insights/2026/08/how-quantity-changes-what-is-possible), and frequently the colour wanted is not available in a small amount at any price. The consequence is that a made to order garment tends to cost most at the moment the brand can least defend a high retail price, because it has no reputation to price against yet.
Cutting can add to the cost difference. A marker that combines several sizes may use cloth more efficiently than a single-garment marker, depending on the pattern, width, print and cutting method. Single pieces can still nest with one another, and some made-to-order operations batch their orders. Compare actual marker consumption and remnant reuse rather than assuming every single-unit cut wastes more cloth.
Setting the line is the same story a second time. Machines are threaded, folders and attachments fitted, the operators taught the sequence, the first pieces checked and corrected. Each restart repeats all of it, and each restart is also a fresh opportunity for the garment to come out slightly different, so the ninth garment is not guaranteed to match the first in the way that the ninth garment off a single run is. Scheduling adds a third effect. A maker fills a line with work that fills a line, so single garments arriving irregularly get fitted between larger runs, and the customer’s wait is set by other people’s orders rather than by the work in the garment.
Trims complete the picture. Labels, care and composition tickets, zips, buttons, polybags and cartons are supplied in packs rather than in units. Either the brand buys packs and holds them, which is stock, or it pays to acquire them in small amounts every time, which is the setup problem again in a different department.
That cost then has to go somewhere. Either it goes into the retail price, which asks a new customer to pay an established brand’s price for an unestablished one, or it is absorbed, which means the brand is buying its own sales. Neither is fatal on its own. Both are worse than they look when the volume is too thin to improve them.
Follow one case through. A founder is selling a single shirt style made to order, and two customers ask for it in a second colour. The colour has to be bought in whatever amount the supplier will sell, which is almost certainly more than two shirts, so the brand now holds cloth. The line has to be set again for the new colour and the first pieces checked again. There is a second photography session, because a colour that is not photographed does not sell. If the colour then moves slowly, the cloth sits. The brand is holding inventory after all, without ever having had the economics of a proper run, and holding it in the least useful form: unsold garments can be photographed, discounted, sent to press, sold at an event or carried into the following year, while unsold cloth still needs a usable design, resale route or later order. Shared stock cloth can be more flexible than a finished garment; custom cloth tied to this style may be much harder to redirect.
## 03The wait is charged at the first purchase, where it is hardest to justify
Lead time is the second cost, and it is charged at the worst point in the relationship. A customer is asked to wait weeks for a garment from a brand they have no experience of, at the exact moment their confidence is lowest. Some will wait. Many will not finish the order, and some of those who do will write to cancel before the cloth is cut. Delivery speed has become part of how an online product is judged, and a long wait reads as risk even when it is presented as craft. The wait is being asked for precisely where it is hardest to justify: the first purchase.
The period between order and dispatch needs deliberate communication. Confirm the expected dispatch window, explain milestones and contact customers promptly if it changes. Watch cancellations against missed promises and the occasions customers bought for. A quiet queue can undermine trust, but it is a service problem to manage, not an inevitable feature of making to order.
There is a worse property hidden in the queue. When several orders arrive together, because a post travelled or a mention landed, the queue lengthens for everyone in it, so the brand’s best week produces its slowest deliveries and its most visible complaints. Success degrades the service. A produced range behaves in the opposite direction, since demand that arrives at once is met from stock that already exists.
Returns complicate this further. Customers still send things back, and a garment cut for one person has nowhere obvious to go when it comes back. Some brands limit discretionary returns, which can raise the barrier at the first purchase. A made-to-order label alone does not establish which customer rights apply; check the rules for the particular product and selling market before writing the policy.
## 04Fit problems arrive one customer at a time instead of in a size set
Fit is where the argument becomes serious. Produced properly, a range meets its fit problems all at once. A graded size set is made, [tried on, measured and argued over in one session](https://surajmalstudio.com/insights/2026/06/how-to-read-a-garment-sample), the pattern is corrected across every size, and the correction happens before bulk. That session is uncomfortable and it is the most valuable hour in the whole development.
What the session actually settles is more than the base size. Measurements are checked against the specification, the garment is seen on bodies at both ends of the range rather than only on the fit model, and the grade rule itself gets tested, which is the part that decides whether the smallest and largest sizes are the same garment or different ones. One revised set of patterns goes back to the maker. Everything produced afterwards inherits the answer.
Made to order does not require skipping a size set. The risk arises when a founder uses the model to postpone fit development until paid orders arrive. Then faults surface on individual customers, and a pattern correction may come only after several garments have been made. An approved block, size-set review and wear testing belong before selling in either production model.
Each of those reports is also an experiment with one participant and no controls. It arrives from a single body with its own proportions and its own idea of how the garment should sit, and there is nothing to compare it against. A pattern fault and a personal preference read identically in a single message. It takes several before the difference becomes legible, which is another way of saying it takes several garments made, paid for and worn before the fault can be named. The observable symptom is worth knowing: two or more customers using the same word about the same part of the garment across separate orders, in different sizes. That is a reason to inspect the pattern, production measurements and size guidance. It is evidence to investigate, not proof of a pattern fault on its own.
The correction costs more here too. A remake is a second full garment plus a return, made under the same conditions of working in ones, so the fault is at its most expensive in exactly the case where the brand is at fault.
## 05One order at a time is too slow and too noisy to design from
Learning follows the same shape. A produced launch delivers a body of evidence quickly: which sizes emptied, which colour stalled, which style carried the others, what people asked before they bought. Made to order releases that evidence in single units, and single units are difficult to read. A quiet month might be the product, the photography, the price, the wait or nothing at all. Decisions get made late and on thin ground.
A concentrated launch can make comparisons easier because products meet customers under similar conditions. It does not establish causation: traffic, stock availability, photography and promotion can still differ. Made-to-order brands can collect demand across a defined window too. Keep those conditions alongside the orders, and avoid redesigning a range from a handful of purchases in either model.
A thin series is also dominated by whoever happened to order. One customer buying two garments at the top of the size range is enough to make the size range look wrong, and one enthusiast ordering three times is enough to make a style look like a bestseller. There is no way to tell an accident from a signal while the total is small, so the size of the evidence matters more than the production label.
This is the deeper cost. A first range is a forcing device. It compels a decision on the size range the brand will stand behind, the fit standard it will hold, the fabric it will commit to, the maker who has to repeat the work and the price the whole thing has to live at. Made to order lets every one of those be postponed to the next order, then the one after. The brand accumulates transactions and no answers, and arrives at the point where it needs to scale with nothing settled underneath it.
A year later that shows up as a specific kind of stall. There is a customer list, a run of transactions and no settled product, so growing means doing the development that was deferred, while carrying customers who are used to the old arrangement. The fit gets corrected in public, on the first produced run, in front of the people who bought the uncorrected version. The price usually has to move as well, because a garment produced properly costs differently from one made in ones, and the direction it moves in has to be explained to people who already paid.
## 06Pre-order collects the demand first and still produces once
What genuinely lowers risk at the start is [a narrower range produced properly](https://surajmalstudio.com/insights/2026/07/against-launching-with-everything): fewer styles, fewer colours, a decided size range, a maker who can make the same garment twice. Pre-order does the same work honestly, and it is a different thing from made to order. Demand is collected first, then everything is produced together in one run, which keeps the production economics, the graded size set and the single fit correction intact while still reducing what is made without a buyer.
The sequence matters more than the label. A sample is made and photographed. Cloth is reserved and a maker booked against a window. The window opens with a stated dispatch date and closes on a stated day. The orders that came in are turned into a size breakdown and scheduled as a coordinated production run, with the necessary markers, lays, line settings and first-piece checks. If the fit needs correcting, it is corrected before the window closes, and the correction reaches every garment.
The customer still waits, but the wait has a different shape: it happens once, it has a date attached, and it is the same date for everybody, so a delay is one message rather than a rolling series of individual apologies. That is also the line between pre-order and made to order wearing a better name. An offer with no closing day and no dispatch date is made to order. The closing day is the thing that turns a set of orders into a production run.
Four decisions have to be made before the window opens, not after. Which sizes will be offered whatever the orders say, because a size range the brand will stand behind is a commitment and not a poll. What quantity the run has to reach for it to go ahead, and what happens if it does not, said in advance and refunded promptly if it comes to that. Whether the window can be reopened, since reopening it repeatedly is how a pre-order quietly becomes made to order again. And what the garment sells for on the day it lands, because a pre-order price set below the eventual retail price teaches the customer the discount lesson before the brand has sold anything at full price at all.
## 07The objection: unsold stock is what kills first brands
It does, and this is the strongest thing that can be said for made to order. A first range that produces the wrong sizes in the wrong colours converts the brand’s cash into goods that will only move at a reduced price, if they move at all. Cash is the constraint that actually ends young businesses, and it ends them regardless of whether the product was good. Made to order attacks that constraint directly and nothing above makes it stop doing so.
The answer is in three parts. The first is that the cash risk is changed in form rather than removed. Cloth bought in the amounts a supplier will sell, trims held in packs, remakes for fit, repeated samples and a long gap between spending and being paid all consume the same cash, and custom cloth and trims can be difficult to redirect. Stock materials shared across styles may retain more options than finished garments.
The second is that what controls stock risk is the size of the bet, not the method of making it. Fewer styles, fewer colours, a size range that has been decided rather than hedged, and a quantity set against demand that was collected instead of assumed: those are the levers, and every one of them is available inside a produced run. Pre-order is simply the version that takes both, cutting the quantity risk without giving up the run.
The third is that the two risks do not destroy the same things. Unsold garments are a damaged asset, and a damaged asset still has uses: photographs, gifts to people who will be seen wearing them, an event, a stockist taken on at a thinner margin, or a following year if the range is not tied to a season. A year spent without a settled fit, a proven maker or a demand signal that can be read is not an asset in any condition. It cannot be sold at a reduction, and the following year has to start it again from the beginning.
Made to order earns its place later, on a settled pattern, with a maker who has already proved repetition: an extended size, a personalised element, a piece whose variation is the reason someone wants it. It works well as an extension of a functioning operation. It is a poor substitute for building one.
Your next step
Put this into practice for your collection.
[Talk about your idea](https://surajmalstudio.com/enquire)
## Read next
### [Build a landed-cost worksheet for a mixed order](https://surajmalstudio.com/guides/what-a-first-order-actually-costs-to-land)
Work through a mixed shipment, assign shared charges and reconcile the estimate with what arrived.
### [Making to order, or holding stock](https://surajmalstudio.com/guides/made-to-order-or-stock)
Making to order and holding stock suit different promises about delivery, different ways of paying for production, and different levels of certainty about what will sell.
### [Packing and shipping a first order](https://surajmalstudio.com/guides/packaging-and-shipping-a-first-order)
Decide how clothes are folded, protected, boxed and identified before packing begins, then use measured boxes and matching records to plan shipping.
### [What can go wrong in a first production order?](https://surajmalstudio.com/guides/what-goes-wrong-in-a-first-order)
First production problems arise when instructions drift, faults repeat unchecked and the final inspection is asked to settle decisions that belong earlier.
Read next [Build a landed-cost worksheet for a mixed order 3 min read · Costs, orders and delivery](https://surajmalstudio.com/guides/what-a-first-order-actually-costs-to-land) [Making to order, or holding stock 7 min read · Costs, orders and delivery](https://surajmalstudio.com/guides/made-to-order-or-stock) [Packing and shipping a first order 7 min read · Costs, orders and delivery](https://surajmalstudio.com/guides/packaging-and-shipping-a-first-order) [What can go wrong in a first production order? 4 min read · Costs, orders and delivery](https://surajmalstudio.com/guides/what-goes-wrong-in-a-first-order)
---
[View this page on the website](https://surajmalstudio.com/insights/2026/08/against-made-to-order-as-a-first-strategy)