For retailers, data is the foundation of forecasting, reporting, and buying decisions. However, when the data has not been cleansed, distorted insights will follow. In fashion, product codes, or SKUs, are one of the most important elements of the business data ecosystem to keep clean, because when one uniform coding system is not upheld, product tracking and analysis chaos is likely.
Product codes form the backbone of accurate product reporting, planning, and analysis. A consistent, clean product code system prevents errors and confusion, making it easier to track inventory, analyze sales, and streamline operations. It also improves communication across teams and with partners. A well-designed coding system will give your entire team a fast, reliable shorthand for what a product is, how it fits into your assortment, and what needs to happen next – from restocking, shipping, or analyzing returns.
So, whether you’re building range plans or reviewing post-season performance, consistency in your product setup is what keeps your decisions accurate and agile. Here’s how to clean your codes to gain actionable insights and help make impactful decisions from their data.
Product code creation and management
What is a product code?
A product code is an alphanumeric code used to track and manage a product. It is not a barcode – a barcode is a standardized numerical code generated for fulfillment, scanning, and external logistics purposes. A product code you create yourself, which is why mistakes and misinformation can occur across teams and warehouses without a properly developed coding system.
Why is product code hygiene important?
Without a clean system for code creation, you risk losing product history, misreported sales, and broken links between styles, sizes, and colorways. Inconsistent product setup leads to broken reporting, distorted sales performance, and misinformed decisions. Platforms like Style Arcade rely on clean product data to function at full power. Without it, users can’t obtain true size curves, accurate sales history, return data, or assortment clarity.
How to create style and colour codes
Step 1: Start with a hierarchy to identify which parts of your business are the most important to break down the code by. Generally, brands will use a department or category as the top-line element of their code creation formula to easily split products and quickly understand what the item is without imagery.
- For example, dresses and bottoms might be coded to 3 characters, such as DRS and BTM
Step 2: Depending on the distribution of your collections, whether seasonal or annual, include the year or season next. The reason for this is to easily identify a product early in its lifecycle, or post, when there is no imagery or name.
- Consider using a ‘season code’ in a numerical format, e.g., AW26 could be 2601, which refers to the year (2026) and quarter (01).
Step 3: Think ahead. Make sure your code has enough characters (e.g., 8-10) to build upon. Your codes will need to stay the same length as you continue to develop your products, and having enough characters will reduce the risk of creating a confusing system.
- For example, instead of DRS123, create DRS000123
- Add +1 when creating a new product to the last, or use a random number generator or formula such as ‘rand’
Step 4: What you’ve created by this stage is known as a ‘style code’. Once you have decided on the colorways in your assortment, you will then need to create a ‘style-color code’.
- To do this, add a “-” then color code, using the same logic as shortening the category to 3 characters e.g., Black = BLK
- Create products in your system as a style-color. Do not set up color as a variant; variants should be left for sizes.
What to do when repeating products
When repeating a product from a previous season to the next, it is important to keep the data history under the same product. You may be tempted to update the product code — avoid this unless your system does not allow price changes over time history.
Instead, use your hierarchy to relabel under the upcoming new season for your reporting.
If you must make a new product, follow the same structure as you normally would. However, you may want to lean on an attribute like ‘Story’ so you can see the history of the first product drop, to the second, over time.
How to treat core codes
- Usually, a product doesn’t start out as core. You don’t need to recode these; just like a repeated product, you could lean on a ‘season’ attribute instead
- If you must, use ‘core’ as a season identifier in the code
Deleting codes
Avoid deleting products once loaded into your ERP/system. Most systems use codes to link data together; you will lose historic sales! (Or not know what they are from).
Archive wherever possible, because when you archive, the product moves out of a visible list, so it feels 'cleaner'. When you delete a product, you may lose sales attached to it, or the sales continue, but you don't know where to 'stick' those sales. When you have no history, how do you learn for the future?
Hierarchy and attribution
The difference between a hierarchy and an attribute
- The hierarchy should be made up of your most important, non-negotiable attributes for a product, e.g., department, category, and season
- Attributes are helpful additional reporting levers; they are like additional pieces of the puzzle to figure out the ‘why’. They are 'optional', generally speaking, when it comes to ERP and product systems. Examples include length, color family, occasion, and subcategory. You don't need these to create a product, but you do need these to group products together for analytic purposes.
How to group data
Keep as topline as possible to allow consistency. Think from a customer perspective, e.g., do they search for ‘dress’ or ‘mini dress’ or ‘red mini dress’. We use these data points to understand the customer and their purchasing behavior, so keeping it close to home is crucial.
- Here are some examples:
- Department e.g. Mens, Kids, Womens OR Apparel, Accessories, Shoes
- Category e.g. Dress, Top, Bottom
- Sub Category e.g. Blouses, Shorts, Skirts
- Season e.g. AW26, SS27
- Length e.g. Midi, Maxi, Cropped
- Sleeve Length e.g. Sleeveless, Short, Long
- Color Family e.g. Black, Red, Pink
- Print e.g. Stripe, Plaid, Floral
- Silhouette e.g. T-Shirt, A Line, Bias
- Occassion e.g. Night, Day, Wedding
- Seasonality e.g. Core, Fashion, Seasonal-Core
How often to review
Ideally, this is a set-and-forget exercise. However, as the brand grows and hits key milestones, you may want to review in line with product strategy moving forward. Additionally, brand data can fall into disarray in the instance that a new planner enters the business and decides to recode the entire brand to suit their previous workflows. Carefully think about whether this is necessary, because data clean-ups are hard work, take a long time, and there is always room for human error, which could cause bigger problems down the road.
Still using spreadsheets?
Platforms like Style Arcade erase hours of product admin, manual data cleansing, and avoid human error from traditional BI tools like spreadsheets, with AI-powered technology that seamlessly integrates your data systems to create one source of truth for key product and sales performance metrics.
With Style Arcade, from trade pack to PDP, fashion brands are unifying metrics and images from all systems into one view, saving reports once and reusing them every time with fresh data, and jumping from high-level analysis to size-level in seconds. Having the same, clean data across every department means clear, actionable insights that drive better product decisions.
