How to Choose Apparel Sizes for a First Small-Batch Run

For a first small-batch apparel run, choose a deliberately broad core size range, then let real customer signals decide where to add depth. Start with the people most likely to buy, check the garment’s actual measurements and fit, and use preorders, sample try-ons or a short sizing survey before committing to every size in equal quantities.

The aim is not to predict the entire market from a spreadsheet. It is to make a sensible first allocation that gives customers a fair chance of finding a good fit without leaving a small brand with boxes of slow-moving stock.

Begin with the product, not a generic size chart

There is no universally reliable split between sizes. A roomy unisex tee, a fitted singlet, a structured jacket and a stretchy legging will attract different customers and behave differently on the body. The first question is therefore not “How many smalls should we make?” but “Who is this garment designed to fit, and how?”

Write a short fit brief before you decide quantities. Include:

whether the garment is intended to be fitted, standard, relaxed or oversized;
the body shapes and proportions you have designed around;
which measurements matter most, such as chest, waist, hip, shoulder or garment length;
whether the fabric stretches, softens, shrinks or holds its shape;
whether customers are likely to layer it over other clothing; and
how much fit variation the design can tolerate before it looks or feels wrong.

Fit language is useful only when it matches the finished garment. Calling a shirt “oversized” does not remove the need for measurements: one customer may want extra width but a regular sleeve, while another may want both width and length. Decide what the ease is doing, then explain it plainly.

Choose the first range you can support properly

A small run does not need every possible size, but it should not quietly exclude the people you have invited to buy. Consider your audience evidence, your construction and the manufacturing constraints together.

For many first launches, a practical approach is to offer a continuous run across the sizes your audience actually uses, rather than producing a single “standard” size and hoping it works for everyone. The exact labels might be Australian numeric sizes, lettered sizes or measurements in centimetres. The labels matter less than consistency and clarity.

Before confirming the range, ask:

What sizes have existing followers, customers or community members requested?
Which sizes appeared in previous sales, if you have comparable products?
Does the pattern grade cleanly across the proposed range?
Will the smallest and largest sizes still have the intended proportions, not merely a scaled-up or scaled-down middle size?
Can your maker produce the full range at the same quality?
Will adding another size create a separate fabric, trim or pattern requirement?

Do not add sizes only to make the range look complete. If a size is offered, it deserves a properly checked pattern, a useful measurement chart and a realistic chance of fitting the intended customer. It can be better to launch with a transparent, tested range and expand quickly than to advertise coverage that has not been properly developed.

Separate the size range from the quantity split

Once the range is selected, decide how many units to make in each size. These are two different decisions. A size can be important to your audience while still receiving fewer units in the first run because demand is uncertain. Equally, a popular size may need more depth, but not so much that it consumes the budget required to test the rest of the range.

Make a simple allocation table with four columns:

size;
planned units;
evidence for the allocation; and
the trigger for a repeat or top-up.

Your evidence might be responses to a survey, sign-ups for a launch list, previous orders, sample try-ons or direct conversations with likely buyers. It does not need to be perfect. Recording why you made the decision is valuable because it stops guesses becoming invisible assumptions.

Keep a small reserve for exchanges if your budget allows. You may need replacement stock in a particular size, especially if the first batch reveals a measurement or grading problem. A reserve is not a reason to overproduce; it is a modest buffer that protects the customer experience while you learn.

Use lightweight demand testing before production

Interest is not the same as a sale, so use more than one signal. A poll can show which colours or sizes attract attention, but a preorder or refundable expression of interest asks customers to make a stronger commitment. You can also invite people from your target audience to try a sample and record both their measurements and their preferred size.

Useful tests include:

A sizing survey: ask for height, usual size, key body measurements, preferred fit and whether the respondent would buy this particular garment.
A try-on session: use a small group with varied bodies, and ask each person to move, sit, layer and describe pressure points rather than simply asking whether they like it.
A preorder window: publish the measurements, fit description, delivery timing and exchange policy before accepting orders.
A sample loan: send a sample to a trusted community member who can report on fit in ordinary use, with clear questions to answer.
A waitlist by size: let people nominate the size they would buy, then separate genuine purchase intent from general enthusiasm.

Ask questions that produce decisions. “What size do you usually wear?” is useful, but not enough: labels vary between makers. Add “What is your chest, waist or hip measurement?” where relevant, and ask whether the customer wants a close, standard or relaxed fit. Make clear how measurements should be taken, and use the same units as the product page.

Sample the ends of the range, not only the middle

A middle-size sample can look excellent while the graded extremes expose problems. Before approving production, check at least one smaller and one larger size, or have the maker provide detailed measurements and construction checks for the ends of the range if physical samples are limited.

Look for changes in proportion rather than just overall width. Check sleeve and body length, neckline position, armhole comfort, pocket placement, rise, waistband stability and the way seams sit when the wearer moves. A garment can technically match a chart and still feel poorly designed if these relationships are lost during grading.

Record the approved sample’s measurements in a fit specification. Include tolerances so you know what variation is acceptable when the run is made. Confirm fabric behaviour after washing or another appropriate care test, because a small change in length or width can alter the customer’s experience.

Sampling is also the point to confirm details that can otherwise cause production trouble. The production planning checklist for trims, labels, grading and quantities is a useful reminder that a sample approval is not the same as being ready to manufacture.

Make your size guide do real work

A size chart should help someone choose, not merely reassure them that one exists. Put body measurements and garment measurements in separate sections. Body measurements describe the customer; garment measurements describe the item laid flat or measured according to your stated method. Mixing the two creates avoidable confusion.

Explain:

where to measure the chest or bust, waist and hip;
whether the measurement should be taken close to the body or over clothing;
which measurement should take priority if a customer sits between sizes;
how much ease the garment includes;
the model’s height and the size worn, if you have permission to publish that information;
the expected length on the body; and
whether the fabric has stretch or the design is intended to be worn oversized.

Use a short fit note beside the chart. For example, explain that the tee is wide through the body but regular through the length, or that the skirt is designed to sit at the natural waist and has limited stretch. Avoid vague phrases such as “true to size” unless you explain what that means for your garment.

Photographs can help, but they do not replace measurements. If possible, show the garment on more than one body and include front, side and back views. Mention any deliberate styling, such as tucking, rolling sleeves or wearing a garment oversized, so customers do not mistake styling for the product’s standard fit.

Plan exchanges before you take orders

A clear exchange process is part of sizing communication. State the time limit, condition requirements, return postage responsibilities and whether a different size is subject to availability. In Australia, also make sure your policy does not attempt to remove rights that apply when a product is faulty, does not match its description or is otherwise not of acceptable quality.

Keep an exchange log during the launch. Record the original size, requested size, reason for the change and whether the customer followed the chart. Patterns matter. Several customers moving up because the upper arm is tight points to a different problem from customers moving down because the body is too wide.

Do not treat every exchange as customer error. It may reveal that the chart is unclear, the grading is inconsistent, the fabric behaved differently than expected or the fit description set the wrong expectation. Use the information to improve the next run and, where necessary, update the current product page.

Protect cash while you learn

Small batches reduce the risk of being stuck with a large quantity of untested stock, but they do not eliminate risk. Manufacturing minimums may apply by style, colour or size, and a low-volume order can have a higher unit cost. Treat the first run as both a product launch and a learning exercise: budget for sampling, measurement checks, packaging, freight, exchanges and a possible correction.

If you cannot afford a deep allocation in every size, choose the smallest responsible test rather than quietly narrowing the range based on habit. Consider a preorder for less certain sizes, a second production window, or a waitlist that tells customers exactly when an additional size may be available. Never promise a restock date you have not confirmed.

Keep colour choices under control if each colour creates its own minimum quantity. A simpler first run can give you more useful information about sizing than a wide colour assortment with too few units in each size. Your product’s job at this stage is to teach you what to make next.

Review the numbers after launch

After orders close, compare planned demand with actual demand by size. Then add exchanges, cancellations, sell-outs and customer questions. A size that sold out first is not automatically the only size you should make more of: it may simply have had a shallow allocation. Likewise, leftover units may reflect poor visibility or confusing information rather than weak demand.

For the next run, review:

units sold at launch and over time;
the number and direction of exchanges;
which measurements customers asked about;
feedback about length, ease and movement;
stock left by size and colour; and
what your waitlist or preorder suggested but did not convert into orders.

Change one thing at a time where possible. If you alter the grading, fabric and fit description together, it will be difficult to know what solved the problem. Keep the original chart and sample notes so you can compare the next version honestly.

A practical first-run checklist

Define the intended fit in plain language.
Choose the range using audience evidence and the garment’s construction.
Test interest by size with a survey, try-on, preorder or waitlist.
Approve measurements and proportions at the ends of the range, not just the middle.
Set quantities separately from the decision about which sizes to offer.
Publish body measurements, garment measurements, fit notes and measuring instructions.
Prepare an exchange process and reserve where the budget permits.
Review sales, exchanges and questions before planning the next batch.

In short, make the first run a measured experiment rather than a guess. Offer a properly tested range, use commitment signals before production, explain the fit with measurements, and keep enough flexibility to respond when real customers show you where the range needs more depth.