Product Definition in MMS: Capturing Retail Detail for Better Analysis

Every transaction a retailer processes, every sales report a category head reviews, and every reorder a store triggers depends on one thing being done correctly first: how the product was defined in the system. Product definition is the stage in a Merchandise Management System (MMS) where a retailer records exactly what a product is, down to the smallest meaningful detail. Get this right, and the business can slice sales data in dozens of useful ways. Get it wrong, and even the most expensive software produces reports nobody can trust. This is why product definition deserves far more attention than it usually receives.

Why product definition is the backbone of retail MMS 🔗

Retail runs on detail. A customer does not buy “a shirt” in the abstract; they buy a full-sleeve, formal, white shirt in size 40. If the system only knows it sold “a shirt,” the buyer learns nothing useful. Product definition is the discipline of capturing those granular attributes so that variants can be analysed separately.

Consider apparel. Two shirts may belong to the same range but differ in sleeve type, collar style, fit, or fabric. These differences drive customer choice and therefore drive sales. The same logic applies to a supermarket. Rice is not one product. It is basmati, sona masoori, and brown rice, each in multiple pack sizes. Jam is mixed fruit, strawberry, and pineapple. Without attributes recorded against each, the retailer cannot tell which flavour or grain is pulling sales and which is sitting idle.

Industry practitioners draw a clear line here: details like sleeve type, wash, and embellishment are best treated as product attributes assigned to items where the difference is relevant, rather than being forced awkwardly into rigid categories. The richer and cleaner these attributes are, the deeper the sales analysis a retailer can perform later.

The business side: merchandise hierarchy and category control 🔗

Before individual products can be defined, they need a home. Retailers organise merchandise into a tiered structure called the merchandise hierarchy, which groups products from broad divisions down to specific items. A common structure moves from the company at the top, through divisions, departments, classes, and finally subclasses that connect to actual items.

The widely cited best practice is to keep this hierarchy to roughly five to seven levels. Going deeper than that tends to add complexity without producing meaningful insight, and it makes top-down planning harder. Most retailers find that five or six levels capture what they need.

Why the hierarchy matters for accountability 🔗

The hierarchy is not just for tidiness. It is the primary mechanism a retailer uses to drive management reporting on sales, inventory, and profitability. Each level becomes a lens for analysis. Instead of looking only at one item, management can view a full breakdown from a department like Menswear right down to a single black-and-white checked shirt.

This structure also assigns responsibility. A category head typically owns the profit and loss for a defined segment, such as Menswear, and is accountable for its performance. Within that, shirt collections may be split by season, like a summer range and an autumn-winter range. When a system associates buyers and merchandisers with hierarchy levels, it gives each person a defined span of control over a product hierarchy. Performance is then tracked against the right owner.

One principle is worth stressing. The hierarchy should reflect how the business is actually managed and how customers actually shop, and it should stay relatively stable year over year. Frequent restructuring breaks the consistency that long-term reporting depends on.

From options to SKU: the language of product attributes 🔗

Within product definition, two terms carry specific meaning, and confusing them causes endless problems later.

An option is a product distinguished by a variant such as colour. A white formal shirt is one option; the same shirt in blue is another. When you add the next dimension, usually size, you arrive at the Stock Keeping Unit (SKU). The SKU is the actual sellable unit that appears on a receipt and in the inventory system. A white shirt in size 38, a white shirt in size 40, and a white shirt in size 42 are three separate SKUs, even though they are one option.

Practitioners describe the SKU as an alphanumeric code that encodes variant details like size and colour, identifying each specific item a customer can buy. Dimensions such as colour and size are what segment inventory into these SKUs, while softer attributes such as sleeve type describe the product without necessarily splitting it into a new sellable unit.

Deciding which attributes lead: the Customer Decision Tree 🔗

How does a retailer decide which attribute sits first in the structure? One common approach uses the Customer Decision Tree (CDT), sometimes called a consumer decision tree. A CDT is a visual roadmap of the steps and criteria a customer considers when buying a product, revealing the hierarchy of attributes that actually matter to them.

The logic is simple but powerful. For ice cream, a shopper might first decide on size, then flavour, then brand. For a kids’ range, the most important split may be age group first, then product type such as toy or dress. By mirroring the customer’s real decision path, the retailer structures its assortment the way customers think, which also makes store layouts easier to shop. Where there is no clear customer logic, retailers fall back on category importance to set the order.

Control points and special rules for SKU management 🔗

Once a product is defined, the MMS lets the retailer attach business rules to it. These control points decide how the SKU behaves across stores and at the till.

Discount privileges are a frequent example. A retailer can specify whether customer or employee discounts apply to a given item. Low-margin products like watches are often flagged so that no further discount is permitted, protecting the slim profit on each sale.

Active and inactive flags keep the system clean. When a SKU is discontinued, marking it inactive removes it from the point-of-sale price list. This trims the list the POS has to search, which speeds up billing at the counter and prevents staff from accidentally selling something that should no longer be on the shelf.

Location listing ensures a product is sent only to stores meant to carry it. Winter jackets need not be listed for a coastal store that never sees cold weather. Season control works alongside this, keeping in-season stock on the floor and pulling out-of-season items out of active selling, which matters greatly for fashion and festival-driven ranges.

Auto-reordering, min-max levels, and delivery scheduling 🔗

Product definition is also where a retailer sets the rules that keep shelves full without human intervention on every item. An auto-reordering flag tells the system to generate a replenishment order automatically once stock drops to a set point.

This relies on min-max reorder levels and a base stock figure for each store shelf. The minimum is the trigger point at which a reorder fires; the maximum caps how much is ordered so the shelf is filled without overstocking. Traditional systems apply fixed reorder points or min-max thresholds at SKU level, though these static rules can struggle when demand spikes suddenly, which is why many retailers are moving toward more dynamic replenishment.

Scheduling deliveries and planning warehouse space 🔗

Defining delivery days by SKU or by vendor helps balance the workload arriving at a distribution centre (DC). If every supplier delivered on the same day, the inbound dock would be overwhelmed. Spreading deliveries across the week smooths the flow of goods inward.

For advanced warehouse management, the retailer can also capture the volumetric space each SKU occupies. Knowing the cubic space a product needs lets the warehouse system run efficient putaway, directing each item to a slot that fits it and using storage capacity well. Without this data, putaway becomes guesswork and space gets wasted.

SKU code, EAN, UPC, and vendor part number 🔗

A defined product needs an identifier the system can use across every transaction. Retailers generally have three options, and a well-built MMS supports all of them.

The first is the retailer’s own SKU code, an internal number used for its private logic. The second is the EAN code, the European Article Number, a 13-digit identifier used by retailers across much of the world to identify products at the POS. The third is the vendor part number, the supplier’s own code, useful when processing purchase orders directly against a vendor’s catalogue.

The EAN and the UPC (Universal Product Code) are closely related. Both are formats of the GS1 Global Trade Item Number, with the UPC being a 12-digit code common in North America and the EAN being its 13-digit counterpart used across Europe, Asia, and most other markets. In practice, FMCG supermarkets lean heavily on the EAN code already printed on a product’s packaging, since it lets the cashier scan and price an item instantly without manual entry. For Indian brands, the choice of format often depends on the target market, as both codes serve the same purpose and most scanners read either.

The unique product code that ties everything together 🔗

At the end of product definition, the system allocates each product a single unique product code. Think of it as a roll number: one permanent identifier that links every transaction touching that product. Purchase orders, goods receipt at the DC, price revisions, sales at the till, and stock counts all reference this one code.

This is what makes end-to-end tracking possible. Because every event points back to the same identifier, the retailer can follow a product from the moment it is ordered to the moment it is sold, and trace its margin, movement, and availability at any point. A well-defined product, with clean attributes and the right codes, is therefore not paperwork. It is the foundation on which every later report and decision stands.

What do you think? If you were defining the hierarchy for a category you know well, such as smartphones or packaged snacks, which attribute would you place first, and would it match how customers actually decide? And how much detail is too much before the extra attributes stop earning their keep in your sales reports?

References

1. https://www.retaildogma.com/merchandise-hierarchy/
2. https://www.toolio.com/post/optimizing-product-hierarchy-and-attributes-for-effective-retail-decision-making
3. https://www.netsuite.com/portal/resource/articles/ecommerce/5-ways-a-merchandise-hierarchy-helps-retailers-increase-efficiency.shtml
4. https://docs.oracle.com/cd/F11448_01/doc.160030/f11690/merchandisehierarchy.htm
5. https://www.relexsolutions.com/resources/consumer-decision-trees/
6. https://algonomy.com/resources/guides/replenishment-optimization-strategies-for-efficient-stock-management/
7. https://www.gs1us.org/upcs-barcodes-prefixes/ean-vs-upc
8. https://support.gs1.org/support/solutions/articles/43000734124-what-is-the-difference-between-a-gs1-gtin-a-barcode-an-ean-and-a-upc-
9. https://www.gs1india.org/blog/ean-vs-upc