How to Start an Online Clothing Store Without Drowning in Returns
An online clothing store has one number that decides everything, and it is not conversion rate. It is returns. Apparel sends more stock back than any other category, and almost all of it comes down to fit. This guide runs the full build: choosing a production model, getting sizing right, controlling SKU sprawl, sampling with a manufacturer, and launching to an audience that already wants the drop.
Don't pay full price to build this.All 39 Ecommerce tools below come with a discount code, a startup programme or free credits.Claim these perks →
Choose the production model before the brand name
There are three production models and the choice sets your cost base for years. Print on demand carries no stock risk and thin margin. Buying wholesale from existing brands gets you saleable product quickly but no differentiation at all. Cut and sew means making your own garments, which is the only route to a genuine brand and the only one that needs real capital.
Print on demand is the right first move for most people, not because it scales but because it teaches you demand cheaply. You learn which designs sell, which sizes people order, and how many come back, all without a pallet of hoodies in the spare room. Treat it as market research that happens to generate a little revenue while it runs.
The migration point is predictable. Once a design sells consistently, print on demand costs you more per year in unit price than a production run would have, and quality complaints start appearing because you control neither the blank nor the print. That is the signal to move to bulk manufacturing, and not a month before it arrives.
Sizing is the product, and returns prove it
Returns decide whether a clothing store works. Apparel has the highest return rate in ecommerce and the overwhelming majority of it is fit rather than fault. Every return costs you the outbound postage, the return postage, the handling time, and frequently the item itself if it comes back worn or the season has already moved past it.
So sizing is product work, not a footer link. Measure your actual garments rather than copying a supplier chart, publish measurements in both centimetres and inches, state the model height and the size worn, and describe fit in plain words: cut close through the chest, generous in the sleeve. Ambiguity on a product page converts directly into a return.
Track returns by SKU and by stated reason from the first month of trading. One style with a wildly higher rate than the rest is almost always a sizing error you can fix with a chart correction or a pattern tweak, and finding it takes a spreadsheet rather than a consultant. Stores that watch only the blended rate never locate the culprit.
SKUs multiply faster than your cash does
A single style in five sizes and three colourways is fifteen SKUs, each needing stock, a barcode, a photograph and a demand forecast. Add two more styles and you are running a warehouse problem instead of a brand. This is where new clothing labels quietly go broke: the money is not gone, it is sitting in the sizes nobody ordered.
Buy the size curve, not equal quantities across the range. Demand across sizes is never flat, and ordering ten of each guarantees leftovers at the extremes and stockouts through the middle where most of your customers live. Use whatever real data you have, from print on demand orders or the first run, and weight the next order towards what genuinely sold.
Keep the range deliberately short at launch. A tight collection with real depth in the core sizes outsells a sprawling catalogue holding one of everything, and it keeps photography, copy and merchandising achievable by one person working evenings. You can always add styles later. You cannot easily unspend the cash locked in the wrong ones.
Tech packs, samples and minimum order quantities
A tech pack is the document that turns an idea into a garment a factory can actually make: graded measurements per size, fabric composition and weight, trims, stitch types, labels and packing detail. Without one you are asking a manufacturer to guess your intent, and they will guess in whichever direction is cheapest and fastest for them.
Sample before you commit to a run. Expect several rounds, because the first sample is almost never right, and the cost of those rounds is trivial set against a full production run of badly fitting jackets. Fit each sample on real bodies at more than one size point, rather than trusting a mid size graded mathematically up and down.
Minimum order quantities are the gate that decides your launch shape. Factories quote per style and often per colourway, which is why small brands launch in two colours and add more later. Domestic and near-shore makers usually accept smaller runs at a higher unit cost, and for a first collection that trade is normally the right one to take.
Photography, product pages and copy that prevents returns
Photography carries the sale because the customer cannot touch the fabric. You need the garment on a person, movement or multiple angles, a close crop showing weave and stitching, and consistent lighting across the entire range so the site does not look like it was assembled from three different brands over three different weekends.
The product page has a job list: fit description, full measurements, fabric and care, model reference, and delivery and returns terms stated plainly rather than buried. Every question you leave unanswered becomes either a lost sale or a return, and the return is by far the more expensive of the two outcomes for a small brand.
Write real copy for every product. Supplier descriptions are duplicated across every store selling the same blank, which is worth nothing in search and nothing in persuasion. Describe how the fabric feels in the hand, where the garment sits on the body, what it works with, and honestly who it is not designed for.
Launch, paid traffic and the numbers that decide it
Launch to an audience, not to an empty website. Build a waitlist while you are still sampling, show the process, and give that list first access to the drop. A collection launched into a warm list sells through its size curve fast enough to fund the reorder, and funding the reorder is the entire game in apparel.
Paid social is where clothing brands scale and also where they burn the reorder money. Start after you know your average order value, return rate and repeat purchase rate, because contribution margin after returns is the only figure that tells you what a customer is worth. New advertiser accounts frequently qualify for starting credits, which softens the learning period.
The rest of the stack is standard and cheap if you claim founder offers instead of paying list price: store platform, design tooling, discounted shipping labels, business banking and the ad credits new accounts qualify for. Our ecommerce directory keeps those together with the live terms, which is worth ten minutes before your first month of subscriptions starts billing.
The offers mentioned above
Pulled live from the directory, so what you see here is what the listing says.
Canva
Design & Creative
$200
CreditsLink on planChatGPT
AI Models & APIs
$5/month
Fixed discountShared accountLink on plan
Mercury
Finance & Accounting
$250 sign-up bonus
OtherLink on plan
Also mentioned on this page with nothing currently on offer: Shopify, Pirate Ship. We track them so you do not have to go looking.
Ecommerce codes, all in one place
39 listings in Ecommerce, with the redemption codes and programme links behind one subscription. Browsing is free.
Common questions
How much does it cost to start an online clothing store?
Print on demand starts at little more than the platform subscription and your design time. Cut and sew is a different order of commitment, because samples and minimum order quantities have to be paid before a single garment sells. Most brands start on the first and graduate to the second.
Is print on demand worth it for a clothing brand?
As a demand test, yes. As a long-term model, it caps your margin and hands quality control to somebody else. Use it to find the designs and sizes that sell, then move those specific winners into a production run.
What return rate should I expect on clothing?
Higher than any other ecommerce category, and driven mainly by fit. Rather than budgeting against an industry average, measure your own rate per SKU from the first month, because the problem is usually concentrated in one or two styles.
Do I need a minimum order quantity to work with a manufacturer?
Nearly always, quoted per style and often per colourway. Domestic and near-shore factories tend to accept smaller runs at a higher unit price, which is the usual compromise for a first collection where you cannot yet forecast the size curve.