A grocery shop selling four hundred products has four hundred things to track. A clothing shop selling four hundred designs might have four thousand, because every design exists in several sizes and several colours, and each of those combinations is a separate thing a customer either can or cannot buy. This is the single reason general retail software so often disappoints garment retailers. The billing works, the reports look fine, and yet the stock figures are quietly meaningless, because the system counts shirts while your customers are asking for a medium in navy.
Table of Contents
Why Clothing Stock Is Harder to Track
Most retail stock behaves predictably. A carton of biscuits is a carton of biscuits, and when the count says twelve, twelve customers can be served. Garments break that assumption, because a count of twelve tells you almost nothing useful. Twelve what? If eleven are extra-large and one is small, the shop is effectively out of stock for the majority of people who walk in asking for that design.
The commercial cost of this is invisible in the accounts. Nobody records the sale that did not happen because the size was missing, and the stock report looks perfectly healthy right up to the end of the season, when what remains is the sizes nobody wanted. Owners often describe this as a buying problem. Usually it is a visibility problem: the buying decisions were made from a report that could not show what was actually selling.
What a Variant Actually Is
A variant is a specific sellable combination of a product’s attributes. One design of kurta available in four sizes and three colours is not one product with twelve units. It is twelve variants, each with its own stock count, its own barcode, and potentially its own price if the extra-large costs more to make.
The distinction matters because it determines what your system can answer. With variant tracking, you can ask which colour sold fastest, which size ran out first, and whether the medium sells better in Lahore than in Karachi. Without it, all twelve combinations collapse into a single number and none of those questions have answers. A garments POS setup is really just ordinary retail software with this one structural difference, and it is the difference that decides whether the data is usable.
Key takeaway: a shop that tracks garments at design level knows how much stock it holds. A shop that tracks at variant level knows what it can actually sell. Only the second one can buy the next season correctly.
Setting Up Size and Colour Without Chaos
Variant setup goes wrong in predictable ways, and almost all of them come from inconsistent naming rather than anything technical. Decide the vocabulary once, before a single product is entered, and hold to it:
- One naming convention for sizes, so S, Small and 38 do not coexist
- A fixed colour list, so navy, Navy Blue and dark blue are one colour
- Attributes defined at product level, then combinations generated from them
- Only combinations you actually stock, rather than every possible pairing
- Design codes that stay stable across seasons for year-on-year comparison
The fourth point saves more time than it looks. Generating all combinations automatically produces a catalogue full of variants you never carry, which clutters every search at the counter and slows billing during a rush. Create what you stock and add the rest when a supplier actually delivers it.
Loading an existing shop’s catalogue is where this becomes real work, since a boutique with six hundred designs is entering several thousand rows. The approach covered in our post on bulk product import applies directly, with one addition specific to garments: sort the spreadsheet by design before importing, so variant groups stay together and errors are easier to spot.
Barcodes and Labels for Garments
Imported and branded clothing usually arrives with a manufacturer barcode already covering the specific size and colour, which means you scan and move on. Locally stitched and unbranded stock, which is most of the market in Pakistan, arrives with nothing, so you generate and print your own labels per variant.
The practical problem is physical rather than technical. Garment labels get detached, curl on hangers, and end up illegible after customers handle them for a week. Tagging at the point of receiving, before stock reaches the floor, is the habit that prevents a staff member guessing at the counter. And a guess at the counter is not a small thing: ring up a medium as a large and you have created two wrong stock figures with one scan, one showing phantom availability and one hiding real stock.
Seasons, Markdowns and Leftover Sizes
Clothing has a clock on it in a way most retail does not. Nothing expires, but a lawn design in October is worth a fraction of what it was in May, and the loss is just as real as spoilage. Variant-level data is what lets you act before that point rather than after.
Broken size runs are the specific thing to watch. A design that has sold out of small and medium but still holds a dozen large is finished commercially, whatever the total count suggests, and it will not improve on its own. Spotting that in week six rather than week fourteen is the difference between a modest discount and a clearance price.
Markdowns are worth planning at variant level too, since discounting an entire design punishes the sizes that were selling perfectly well. Handling exchanges properly matters here as well, because size swaps are constant in clothing and each one moves two variants at once; the discipline in our post on returns and exchanges keeps those figures straight.
Reports That Only Make Sense at Variant Level
Once variants are set up properly, a set of reports becomes available that simply cannot exist without them: size curve by design, colour performance across a collection, sell-through rate per variant, and broken size runs flagged before they become dead stock.
The size curve is the one worth building a buying routine around. Every shop has a distribution that reflects its own customers, and it is rarely the even split suppliers push. A store that knows it sells medium and large at twice the rate of small can order to that shape instead of accepting a standard ratio and absorbing the leftovers season after season. Pulled from a decent reporting dashboard, that single figure changes purchasing more than any other report in a clothing business.
Conclusion
Garment retail is not harder than other retail, but it is structurally different, and software that ignores the difference will produce confident numbers that mean nothing. Set up sizes and colours with a consistent vocabulary, tag stock at receiving, watch broken size runs weekly, and let the size curve drive your buying. Everything else in a clothing shop gets easier once the variant layer is right.
Running a clothing store on software that cannot see sizes?
We will set up your designs, sizes and colours properly and show you what your size curve actually looks like.
myPOS works with clothing and garment retailers across Pakistan, and the retail management system handles variant-level stock as standard rather than as an add-on.
Frequently Asked Questions
What is a variant in a clothing POS?
A specific sellable combination of a design’s attributes, such as a particular kurta in medium and navy. Each variant carries its own stock count and barcode.
Do I need a separate barcode for every size and colour?
Yes. Without a unique code per variant, the counter cannot record which specific item was sold, and the stock figures stop reflecting what is on the rail.
How do I handle unstitched fabric sold by the metre?
Fabric is tracked by unit of measure rather than size, with colour or design as the variant attribute and quantity recorded in metres or yards.
Can a clothing POS handle exchanges for a different size?
It should. A size exchange returns one variant to stock and removes another, and both movements need recording against the original invoice.
Is variant tracking worth it for a small boutique?
Usually yes, because small shops feel a broken size run more sharply. Even fifty designs across three sizes is more combinations than anyone tracks reliably by memory.
