The verdict in three sentences
A Lagos fashion store listing 80 styles in 3 sizes and 4 colors does not manage 80 products but 960 distinct SKUs. Without a variant matrix with per-SKU stock, phantom stock-outs and oversells explode: 13% of orders cancelled in 2026. Structuring variants brings that down to 3% and cuts publishing time by nearly four.
Single product vs variant matrix
The classic trap: creating one "Ankara dress" product with a global stock of 30, when real stock is spread across 12 size/color combinations. The customer orders a red M that is out of stock because the system only counted the total.
| Criterion | Single product (global stock) | Variant matrix (per-SKU stock) |
|---|---|---|
| SKUs actually managed | 80 | 960 SKUs |
| Orders cancelled for stock-out | 13% | 3% |
| Publishing time per product | 22 min | 6 min |
| Real stock visibility | None per combination | Per size and color |
| Oversells on best-sellers | Frequent | Rare |
| Targeted restock | Impossible | Per exact SKU |
The recommended data model
The right structure separates the parent product from its child variants, each with its own SKU, stock and possibly price.
| Field | Parent product level | Variant (SKU) level |
|---|---|---|
| Name / description | Yes | Inherited |
| Images | Yes + per color | Per color |
| Price | Base price | Adjustable (large size +X) |
| Stock | Computed sum | Real per SKU |
| SKU code | Root (e.g. DRESS-ANK) | DRESS-ANK-M-RED |
| Alert threshold | Global | Per SKU |
| Available status | Auto if any SKU in stock | Per SKU |
2026 rule: the stock shown to the customer must always be that of the selected variant, never the total. A per-SKU alert threshold triggers restock before popular combinations run out.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Mini case study
Amaka runs a fashion store in Lagos, 80 styles, 45,000 FCFA average basket, 200 orders a month. With global stock, 13% are cancelled for phantom stock-outs, i.e. 26 lost orders, i.e. 1,170,000 FCFA evaporating monthly. She moves to a variant matrix: the rate drops to 3%, i.e. 6 cancelled orders, so she saves 20 orders, i.e. 900,000 FCFA recovered monthly. As a bonus, she gains 16 minutes per product at publishing, over 21 hours saved across her catalog.
FAQ
Why do my best-sellers always run out? Because global stock hides which size/color combination sells fast. With per-SKU stock, you see the red M moves 5x faster and restock it in time.
How many SKUs is that really? Multiply styles x sizes x colors. 80 styles in 3 sizes and 4 colors = 960 SKUs. That number, not the 80 products, must drive your stock management.
Does a variant matrix slow down publishing? The opposite: with a parent product and automatic SKU generation, you go from 22 to about 6 minutes per product. Manual entry combination by combination is what takes time.
Do I need a different price per variant? Often yes for large sizes or special fabrics. The parent/variant model allows a base price adjustable per SKU without duplicating the listing.
What cancellation rate should I target in 2026? Below 3% of orders cancelled for stock-outs. Above 10%, you lose both revenue and customer trust, and a variant-structure audit is due.
Let's talk about your project. We structure your fashion catalog into a variant matrix with per-SKU stock and restock alerts. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
