What strains the data model in textiles
The strain comes from variant explosion: a single style becomes hundreds of stock records through a colour × size matrix. When those variants are opened as separate product records, order, production and stock reports can no longer be read at style level. The correct model is a variant hierarchy beneath a parent style.
Critical processes
- Colour and size matrix: variant management under a style, with reporting at both style and variant level.
- Collection and season: style lifecycle, seasonal pricing and markdown planning.
- Sampling: sample requests, revision rounds, customer approval and the approved sample as the production reference.
- Subcontractor tracking: cutting, sewing, washing, printing and embroidery sent out, with quantities dispatched, returned and lost.
- Fabric consumption: theoretical consumption per style, marker plans, actual consumption and variance analysis.
- Quality and seconds: managing defective goods as separate stock.
Which solution maps to what
Style and variant structures, bills of material, subcontracting and costing fall under [ERP consulting and development](/en/solutions/erp-consulting). Wholesale channels use the [B2B dealer portal](/en/products/b2b-dealer-portal), own stores use [retail POS](/en/products/retail-pos), online channels use [e-commerce software](/en/products/ecommerce-software), and workshop terminals use [mobile app development](/en/solutions/mobile-app-development).
Mandatory integrations
Cutting room and sewing line terminals, barcode and label printers, e-invoicing and e-dispatch, marketplace channels, shipping carriers, bank statements and accounting.
Where we start
In textiles and apparel the starting point is usually a variant explosion: one style in 8 colours × 6 sizes becomes 48 lines, and those lines are opened as separate products in the stock system. Style-level reporting becomes impossible and collection performance stays invisible.
The first exercise produces a style-variant hierarchy: the style is the parent record, colour and size are its dimensions. With that hierarchy, ordering and reporting are both managed from a single line; without it, every season grows the database without improving visibility.
What the data model must make explicit
- Style and variant: style code as parent; colour, size and where relevant print as dimensions. Barcodes are generated at variant level.
- Size runs: run definitions that differ by customer or market (S-M-L and 36-38-40 on the same product).
- Collection and season: season, collection and selling period bound to the style record.
- Subcontracted operations: the externally performed parts of cutting, sewing, washing and pressing, with quantities sent, returned and scrapped.
- Fabric consumption: the variance between theoretical consumption (marker) and actual consumption per style.
- Quality grades: first quality, second quality and defective stock held separately.
Roll-out sequence
1. Style structure: style-variant hierarchy, size runs and barcode logic.
2. Production and subcontracting: work orders, send/return tracking, fabric consumption and scrap recording.
3. Sales channels: wholesale ordering, the [B2B portal](/en/products/b2b-dealer-portal), store and online stock synchronisation.
4. Costing and collection analysis: actual cost per style and end-of-season sell-through.
What we measure
- Sales distribution by size and colour (how well the production run matched demand)
- Variance between theoretical and actual fabric consumption
- Quantity difference sent to and returned from subcontractors, and scrap rate per subcontractor
- Style and variant distribution of stock remaining at season end
- Share of second quality and defective goods
Common mistakes
- Opening variants as separate products. Style-level analysis becomes impossible and the item list grows unmanageable.
- Forcing a single size run standard. Different markets use different runs; one run means manual conversion at order entry.
- Posting subcontractor scrap to general waste. Which subcontractor produces how much scrap stays invisible.
- Keeping second quality in the same stock record. Sellable stock looks larger than it is.
Frequently asked questions
Should variants be separate product records?
No. The style should be the parent record with colour and size as variants. A flat structure forces manual consolidation for collection performance and style-level sales.
How is loss tracked in subcontracted work?
Quantities dispatched and returned must be recorded separately and the difference compared against the contractual tolerance. Without that tolerance in the system, every subcontractor account is settled by hand.
Can marker planning be connected to the ERP?
Marker software usually runs separately; the connection is made by transferring the theoretical consumption value into the ERP and comparing it with actual. Without that comparison, fabric waste cannot be measured.
Should barcodes be at style or variant level?
At variant level. Reusing the same barcode across sizes or colours produces differences at the till and during counts that cannot be separated afterwards. The style code is kept as its own field and reporting runs on it.
How is end-of-season stock handled?
Remaining stock must be visible by style, colour and size; a total quantity drives no decision. Outlet, markdown or bundle decisions rest on that breakdown. If the season field is not bound to the style record, the analysis is redone by hand every time.
Is a separate portal needed for subcontractor tracking?
Not necessarily, but as the number of subcontractors grows, chasing them by e-mail and phone lengthens waiting times. The [B2B portal](/en/products/b2b-dealer-portal) infrastructure can also let subcontractors report dispatch and return; what matters is that the report matches a work order step.