Before a company can rationalize SKUs, it must understand the portfolio in a structured way. Most businesses begin with a flat item list from an ERP or planning system. That list is necessary, but it does not explain how products relate to one another, where complexity is concentrated, which items are genuinely core, or which ones survive mostly because no one has challenged them. Rationalization decisions made from a flat list usually become blunt.
This chapter turns the item list into a portfolio view. It explains how to define a usable SKU hierarchy, how to segment the assortment so patterns become visible, how to separate the base business from the tail, and how to classify items by lifecycle stage, strategic role, and substitution behavior.
5.1 Defining the SKU Hierarchy and Product Taxonomy
The first discipline is to stop treating every SKU as a stand-alone object. Most portfolios have a natural structure. A brand contains families, families contain variants, and variants appear in different sizes, packs, colors, or configurations. In industrial settings, a platform may contain models, options, accessories, and replacement parts. When that structure is unclear, companies compare items that should be grouped together and group items that should be treated separately. The result is confusion about where complexity really lives.
SKU hierarchy: the relationship between the most detailed item level and the broader product levels above it. A good hierarchy lets the business roll performance up and down the portfolio without losing meaning. That is essential because rationalization decisions are rarely made at only one level. A company may keep a family but reduce pack sizes within it. It may preserve a model but eliminate redundant configurations. It may support a style while cutting colors that duplicate demand. Without a clear hierarchy, those choices are hard to make consistently.
Product taxonomy: the standardized set of categories and attributes used to classify products across the business. Typical attributes include size, format, flavor, material, dosage, fitment, pack type, channel, geography, and status. The value of taxonomy is comparability. Instead of relying on inconsistent description text, the business can analyze structured attributes and see where unnecessary variety has accumulated.
In practice, taxonomy work is often less glamorous than analysis, but more important. Systems rarely use one naming convention. One file may say “12 oz,” another “12OZ,” and a third “0.75 lb.” Some items are classified by brand, others by plant, others by historical launch code. Before the portfolio can be understood, these inconsistencies have to be normalized. The question is not simply how many SKUs exist. The question is how many meaningful variants exist once the noise is removed.
Parent-child relationships also matter. Bundles, multipacks, displays, service kits, and customer-specific labels can distort the view if they are mixed together without logic. Some should be analyzed as separate SKUs because they create distinct demand and handling requirements. Others should be linked to an underlying base item so the business can see that what looks like many products is really one product wrapped in several forms. The goal is practical clarity: a structure that shows where complexity sits and at what level decisions should be made.
5.2 Segmenting SKUs by Category, Channel, Customer, and Region
Once the hierarchy is in place, the next step is segmentation. A SKU may look weak in the total portfolio and still be essential in one route to market. Another may appear important because of total sales, yet depend almost entirely on one customer. Segmentation reveals those patterns. Without it, rationalization teams either overprotect items because they can imagine a possible use somewhere, or overcut items because the real context is hidden inside an aggregate number.
Segmentation: grouping SKUs according to the commercial or operational context in which they are sold and served. The most useful dimensions are usually category, channel, customer, and region. These are not just reporting cuts. They explain why an item exists and what would happen if it changed.
Category segmentation helps determine whether a SKU is part of a coherent range or an isolated exception. Within the same broad category, assortment logic may vary significantly. A premium subcategory may justify more breadth than a value tier. A technical subcategory may require real differentiation, while another is overspecified and full of near-duplicates. Category segmentation prevents the business from applying one simplification rule across fundamentally different product groups.
Channel segmentation matters because routes to market reward different types of variety. A pack that works well in e-commerce may not deserve shelf space in mass retail. A club-size offer may be valid wholesale and inefficient everywhere else. A direct-sales channel may tolerate more customization than a distributor network. If channel context is ignored, the company either keeps too much channel-specific clutter or cuts items that are actually justified by that channel.
Customer segmentation usually reveals where hidden complexity sits. Many tail SKUs are effectively custom items for one or two accounts. Some earn their place because they support a strategic relationship, a large basket, or a binding commitment. Others survive because a one-time request was never revisited. Segmenting the portfolio by customer dependence allows the business to separate strategic exceptions from historical accidents.
Regional segmentation adds a final layer of context. Preferences, regulatory requirements, pack norms, and competitive conditions differ by geography. What looks like duplication from headquarters may be necessary locally. At the same time, regional views often expose avoidable divergence. The aim is to tag each SKU with the few context flags needed to explain whether it is broadly core, locally justified, or unnecessarily fragmented.
5.3 Distinguishing Base Business From Tail SKUs
Every portfolio has a base business and a tail. The base is the set of SKUs that carries most of the volume, revenue, and usually margin. The tail is the collection of items that individually contribute little, sell irregularly, or create more effort than their value justifies. Rationalization commonly starts in the tail, but many teams define the tail too mechanically.
Base business: the core group of SKUs that forms the stable economic foundation of the portfolio. Tail SKUs: items with low velocity, narrow demand, or limited economic contribution relative to the complexity they create. Volume matters, but it is only one part of the distinction.
A better way to separate base from tail combines several signals. Sales concentration is the obvious starting point: in many portfolios, a small share of SKUs drives most of the business. But two low-volume items can have very different profiles. One may be stable, profitable, and easy to serve. Another may sell unpredictably, require dedicated components, tie up safety stock, and create write-offs when forecast incorrectly. Tail assessment should therefore look at demand stability, margin quality, account concentration, inventory health, and operational burden in addition to topline sales.
It is also useful to distinguish between a healthy niche and an unhealthy tail. A healthy niche SKU serves a real, defensible need and is managed intentionally. Its volume may be modest, but its role is clear. An unhealthy tail SKU lacks that clarity. Substitutes exist, the demand pattern is weak or sporadic, and the item remains active mostly because no one has forced a decision. Rationalization should focus first on an unhealthy tail rather than treating every small SKU as a problem.
Tail status is also dynamic. A new launch may look like a tail item before distribution builds. A seasonal product may appear weak outside its peak window. A declining item may still be active only because inventory is being run down or a customer transition is underway. This is why base-versus-tail classification should use a relevant time horizon and consider seasonality, launch timing, and end-of-life timing. When done well, the distinction becomes a management tool.
5.4 Identifying Lifecycle Stage, Strategic Role, and Substitution Patterns
The final layer of portfolio understanding goes beyond performance and asks three practical questions. Where is this SKU in its life? Why is it important, if it is important at all? And where would demand go if the item changed or disappeared? These questions help explain why two SKUs with similar sales can deserve very different decisions.
Lifecycle stage: the point a SKU occupies in its commercial life, such as launch, growth, maturity, decline, seasonal return, or end-of-life. The same sales profile means different things at different stages. A recently launched item may be small because it has not yet reached distribution. A mature core item may grow slowly but still anchor a large repeat-purchase base. A declining SKU may still appear active only because the business is running down stock or managing a phased customer migration. Lifecycle context prevents the business from cutting the future or preserving the past for the wrong reasons.
Strategic role: the reason a SKU matters beyond its direct sales. Some items are entry points into the brand. Some defend a price tier or complete a line for retailer credibility. Some support an installed base, a regulatory requirement, or a technical standard that cannot be abandoned casually. Others generate halo value by signaling innovation or breadth. Strategic role should not become a blanket excuse for keeping everything, but it should be identified explicitly so that the portfolio is judged fairly.
Substitution pattern: the likely path by which demand would move if a SKU were simplified, replaced, or discontinued. This is one of the most important concepts in rationalization because value is rarely lost one-for-one when an item exits. Customers often shift to the nearest acceptable alternative. The portfolio must therefore be read as a network of overlapping options, not just a set of disconnected item numbers. Which products are true substitutes? Which are only partial substitutes? Which requires communication, conversion, or technical validation? Which surviving SKUs can absorb transferred demand without creating new service or capacity problems?
Good substitution mapping is both commercial and operational. Commercially, the replacement has to make sense to the buyer or user. Operationally, the surviving item must be available, scalable, and simple enough to support. In some cases, the best move is not simply to delete a weak SKU. It is to guide customers toward a better, clearer alternative and redesign the range so the migration feels logical. Once the lifecycle stage, strategic role, and substitution are visible, the portfolio stops looking like a cluttered item file and starts looking like a system that can be managed with intent.
