The launch date is a design constraint

A launch date is a design decision. It is usually taken as a marketing one, chosen because a season looks right or because a certain amount of time feels reasonable, and by the time the range reaches the table the date has already ruled things out.

01The date fixes the weather the clothes are bought for

The first thing it settles is the weather the clothes are bought for. A customer shops for the conditions in front of them, not for the ones the collection was drawn in. Landing a first range in one part of the year sets it against a particular kind of wear, and that decides cloth weight, fibre, whether a garment is meant to layer, how much lining it carries and how warm it is allowed to be. Move the date across a season boundary and a large part of that stops being a choice. It has been made.

What follows from that is larger than a fabric weight. The conditions decide which garment in the range is the one people arrive for, and therefore what the brand looks like from outside on the day it opens. A range that lands in the cold is read through its outer layers whatever else is in it, because that is the garment a customer is shopping for at the moment they find the brand. The same clothes launched into warm weather are read through the lightest thing on the page. The founder has an idea of what the brand is about; the season decides which part of that idea gets tested first, and the first impression is the one that gets repeated back.

There is a quieter version of the same trap in the drawing. Ranges tend to be designed in the conditions the person designing them is living in, because those are the clothes they can picture wanting. A collection drawn in one season and sold into another arrives full of small assumptions that were never checked: a sleeve length that felt generous at the time, a cloth chosen because it felt good in the hand in a cold room, a colour that read as rich under one kind of daylight and reads as heavy under another. The date is what pulls those assumptions into the open, but only if it is stated early enough for the range to be drawn against it rather than adjusted to it.

It also explains why a slipped date costs more than the slip. Clothes do not move with the calendar. A garment resolved for one set of conditions and delivered into a different set is not late, it is wrong, and the choice at that point is between waiting for the conditions to come round again or re-cutting the range so it suits the ones that have arrived. Either is expensive, and neither is what anybody had in mind when the date moved by what looked like a small amount.

02Rounds of sampling are how a fit gets corrected, and the date sets how many there are

The second thing the date settles is how many rounds of sampling fit before production has to begin. Each round is one opportunity to correct something real: a shoulder that pulls, a hem that flutes, a pocket bag showing through a light cloth, a sleeve that is right on the stand and wrong on a person. Rounds are the mechanism by which a fit becomes correct. The date sets how many are available, and that number tests whether the proposed range can be resolved in time. If it cannot, reduce the scope or move the date; unresolved safety or performance requirements are not a calendar trade-off.

Rounds are serial, and that is the part founders are surprised by. Corrections cannot be run alongside each other and added up, because they act on one another. Taking width out of a shoulder changes where the sleeve head sits, which changes how the sleeve hangs, which changes where the cuff falls on a wrist. Lifting a neckline changes the balance of the whole front and can move a hem that was level. Every change has to be seen on a body against the change before it, which is why two rounds are not one round done twice as quickly, and why a schedule cannot buy back a round by working harder inside the one it has.

A round is also more than a garment being sewn again. It is the pattern being altered, the pieces cut, the garment made, sent, fitted on a person who is not the founder, looked at from behind as well as in front, photographed, and written up so that the next maker of the same garment reads the same instruction. Take a round out and a specific fault stays in: a sleeve head slightly too full for the armhole, a neckline that gapes when someone leans forward, a back rise that is comfortable standing and tight sitting down. None of those stop a garment being shipped. All of them are the reason it comes back.

The rounds are not interchangeable either. Fitting in a stand-in cloth answers questions about proportion and little else, because drape, recovery and the way a seam settles all belong to the real material. The round that matters most is the one done in the cloth the customer will actually receive, after it has been washed the way the label will tell them to wash it. That is also the last round in the sequence, which makes it the first casualty of a tight calendar. The schedule does not announce that it is removing the most informative fitting. It simply runs out at the point where that fitting would have been.

03A cloth that cannot arrive in time is not a choice

The third is which fabrics can be got at all. Cloth is made in runs and dyed in batches, and a mill schedules its capacity around its own programme. Its calendar is not a negotiation, and it takes no interest in which week a brand has chosen to open. A cloth that suits the idea may be unreachable inside the window the launch has created, and the range then gets built from what was available. That substitution rarely shows in a photograph. It shows in the garment, in how it hangs and in how it behaves after washing.

The mechanism is worth being plain about, because it is the one part of the schedule that no amount of urgency touches. A cloth held in stock exists now and is available to everybody who wants it, which means it is available until it is not. A cloth woven or knitted for an order exists when the mill has a slot for it between the programmes it has already committed to, and a colour dyed for an order exists when the dyehouse runs the batch it belongs to. None of that responds to a brand explaining why its date matters. The only real lever is asking earlier, which is a decision about the calendar rather than about the cloth.

Substitution is where the cost lands, and it lands in the place with the least room left. A heavier or lighter weight changes the drape, which reopens the fit that has already been signed off. A different construction takes up and shrinks differently, which reopens the measurements and therefore the size guidance. A different fibre blend takes dye differently, so a colour approved on the first cloth is not the colour that arrives on the second. Each of those is a return journey through work that was supposed to be finished, taken at the point in the calendar where there is no time to make it. The observable symptom is a range that gets flatter and safer as the date gets closer, not because anyone chose safety, but because the interesting options were the ones that could not be delivered.

04The photographs, the site and the words all wait on a finished sample

A launch is not one thing arriving. It is a set of things that all take the finished garment as their input. The photographs need it on a person. The website needs the pictures, and the pages need measurements taken off a made garment rather than off a pattern. The colour has to be named after it has been seen. The description of what a garment feels like and how it moves cannot be written from a drawing, and the size guidance cannot be written at all until somebody has worn it. Packaging has to fit the actual folded bulk of the thing. Anyone selling it on a brand's behalf needs the same material earlier still.

Because all of that sits downstream of one object, a late sample compresses every piece of launch work at once, and the compression takes a predictable form. The shoot gets booked against the date rather than against the garment, so it happens on whatever exists when the studio is booked, which is the sample. The sample was made by the most experienced hands in the building, in the first cloth that arrived, and finished with attention that a run of garments will not receive. The pictures are then true of the sample and slightly untrue of the parcel, and the difference shows in the two places customers look hardest: the colour and the length.

That gap is the most expensive kind of return, because the customer is right. A garment that did not fit is a fitting problem and can be answered with better size guidance. A garment that does not match its photograph is a trust problem, and the answer to it is a refund and a sentence in a review that stays legible long after the range has sold through. Publishing measurements taken from a pattern instead of a garment has the same shape: the number is not a lie when it is written, but it is what the parcel gets measured against, and it is the brand that is judged wrong when the two disagree.

05Set the date from the longest lead item, and work out which one that is

So the date should be set backwards, from the item with the longest lead. That is usually fabric, because it sits in front of everything else: nothing is cut, sampled in the real cloth, graded or approved in bulk until the material exists. Work back from it and the launch window becomes an outcome of the range. Work forwards from a wish and the range becomes an outcome of the window.

Finding the longest lead item is a mechanical exercise rather than a judgement call. Write down everything that has to exist on the day the brand opens. For each of those, write down what has to exist before it can be made. Keep going until each chain ends at something nobody in the brand controls, and then compare the chains. The chain with the longest combined duration and dependencies sets the earliest credible date, with allowance for uncertainty, and it is almost never the thing that feels most urgent day to day. Work the studio controls can be pushed by working harder at it. Work that sits in somebody else's production programme needs their confirmed capacity and cannot be shortened by assumption, and the schedule is only as honest as its longest chain.

It is worth naming that item out loud rather than assuming it is the cloth. It might be a trim that comes from one source and has to be approved before anything can be cut. It might be a wash or a specialist finish, which is a separate building with its own queue. It might be a woven label, a hardware component, or a piece of packaging that has to be made to a size the garment has not been confirmed at yet. The point of naming it is that it becomes the thing the calendar is built from, and the thing whose delay is treated as a real event rather than as a detail somebody will chase.

If the brand is selling to shops as well as direct, the arithmetic runs twice, because a stockist buys and takes delivery ahead of the customer, and the date that governs the clothes is the earlier one. This catches founders who plan a direct launch, build a sensible calendar for it, and then add a wholesale conversation partway through. The buying happens against a calendar that was set by somebody else long before the brand existed, and the range has to be shown in a state good enough to sell from well before it is finished. A schedule built for one route does not extend to cover the other; it has to be rebuilt from the earlier of the two dates.

06Compression comes off the end, so it is the fit and the colour that get cut

Compression is never spread evenly across the work, which is what makes a short calendar dangerous in a specific way. The beginning of a project takes whatever time it takes: deciding what the brand is, drawing, changing your mind about a silhouette, arguing about the one garment the whole thing rests on. Nobody shortens that, because it is happening before anyone is counting. The time comes off the end instead, out of fitting, colour approval and the checks before bulk, which is precisely the part that determines whether the clothes are good. A late start does not produce a rushed idea. It produces a rushed garment.

The reason is structural rather than a failure of discipline. The early work has no wall in front of it. Nothing forces the decision about a silhouette to be finished on a particular day, so it finishes when it finishes, and the time it took is only visible afterwards. The late work sits directly against a fixed production start, and a fixed production start is the only hard edge in the whole calendar. Everything that overruns upstream arrives as pressure on the one stretch of work that cannot expand. A schedule that looked generous at the beginning has already spent its fit rounds by the time anybody looks at it.

What that looks like from inside is a change in how approvals get given rather than an announcement that corners are being cut. Sign-off moves from garments to photographs of garments, because a photograph arrives faster. Colour gets approved on a screen instead of against a physical swatch under a consistent light, which means it is being approved under whatever that screen is doing in that room. Comments get shorter. A fault gets noticed, discussed, and left, because the only available response would cost a round that no longer exists. Nobody decides to accept a lower standard. The calendar decides it for them, one small accommodation at a time.

This is why a date set without the backwards arithmetic does not usually get missed. It gets paid for. When the calendar tightens, the date is the least likely thing to move, because it has been announced, booked and told to other people. Everything softer gives way around it.

The payment comes out of the product. A second-choice cloth, because the first could not be delivered in time. One fit round fewer than the garment needed, so a pattern goes into production carrying a fault everybody noticed and nobody had time to correct. A colour signed off quickly, under whatever light was in the room. None of that registers as a missed deadline. It registers later, as a return rate, as photographs that cannot be used, as a garment the founder quietly stops wearing.

Those consequences outlast the launch they came from, which is the part that is hardest to see at the time. Returns are not a one-off cost. A fit fault comes back as the same size, repeatedly, and the sensible-looking response is to buy less of that size next time, which turns a pattern problem into a permanent buying decision and quietly shrinks the range of people the brand fits. Reviews are written once and read for as long as the product page exists. And the data stops being readable: with a fault in the pattern, a founder cannot tell whether a garment sold badly because people did not want it or because the ones who wanted it could not wear it, so the next range is designed from a signal that was never clean.

The shape of it is clearest in a worked case. A founder sets a launch to a selling moment because that moment looks like the busiest point in the year. The date is fixed first, so the fabric has to come from what can be got inside the window, and the range arrives a weight lighter and a shade more ordinary than the one that was drawn. The remaining calendar has room for a first fitting and one correction, and the correction is spent on the garment the founder cares most about, so the others go into production as they are. Photography is booked against the date rather than the samples, so the shoot happens on samples, and the pictures show a colour the bulk does not match. The range that opens is a different range from the one that was designed, and every one of those differences was decided by the date rather than by anybody in the room.

07When the date is genuinely worth more than the range

A useful test is to ask what the date is for. If it exists because one selling moment is genuinely worth more than the others in the year, that is a real reason, and it is worth designing towards even at some cost. If it exists because it seemed like a sensible amount of time to allow, it is a guess wearing the clothes of a decision, and the range will be the thing that enforces it.

There are dates that genuinely cannot move. Money that is only available inside a window. A stockist working to a buying calendar that was set without reference to anyone else. A collaboration whose other side has committed to its own timing. In each of those the date is a real constraint and should be treated as one, but designing towards a fixed date means designing towards it, not compressing the checks and hoping. The honest version is a smaller range: fewer styles, fewer colours, cloths that are held in stock rather than made to order, a construction with fewer unknowns in it. That range can be resolved inside the window. The original range cannot, and shipping it unresolved is not the same decision as shipping less of it well.

The strongest objection to all of this is that a brand without a deadline never ships, and that objection is correct. Without a date, fit rounds run indefinitely, because there is always one more thing that could be improved and improving it is more comfortable than putting clothes in front of people who might not buy them. A range can be refined until the season it was made for has gone. The answer is not to soften the date. It is to move the flexibility somewhere it does less harm. A fixed date with a fixed range forces the compression onto the fit and the colour, because those are the only things left that can give. A fixed date with a range that is allowed to shrink keeps the same discipline and pays for it in styles instead, and styles can be added to a brand that already fits well. A fit fault cannot be removed from garments that have already been sold.

Which is why moving a launch is usually cheaper than shipping a range that was not ready. Moving costs a rescheduled shoot, a delayed announcement, an awkward conversation with everyone who was waiting, and possibly a season of selling. Shipping early costs the first impression, and the first impression is made once. The earliest customers are the ones who talk about a new brand, and clothes that fit badly give them the wrong thing to say.

The decision is also far easier if it is taken early. Early on, a launch date is a plan, and moving a plan costs a conversation. Later, the same date is a set of commitments: photography booked, stock ordered, delivery agreed with a stockist, pre-orders taken and paid for. Moving those costs money, and it costs standing with the people who arranged their own work around the date. The evidence that the timing is wrong almost always arrives well before anyone is willing to act on it, and the gap between knowing and saying so is where most of the damage happens.

Set the date with the fabric in front of you, and argue about it as seriously as you argue about the clothes. It is going to make choices about them either way.